DE602004004246T2 - Method and system for detecting defects and dangerous properties of passing railway vehicles - Google Patents

Description

1 technical area

These The invention relates to safety in rail transport, and in particular to the detection of at least one faulty one and / or dangerous Conditions for a train formation of at least one passing rail vehicle, the dangers of the gauge line profile, slipping loads, overheating, losses and incipient failures in the axle bearings, overheating the wheels and brakes, overheating of the vehicle body and fire on board. More specifically, it concerns a procedure and a system with which one can make a series of recognition functions for Can cause defects and dangerous situations of rail vehicles, by placing sensors and measuring instruments along the rails. There are also some aspects such as the maintenance of rolling stock, Maintenance of rails and rail transport logistics related to the integration of the system according to this Invention with other railway systems and / or systems with reference to addressed the web.

2 State of the art

2.1 Introduction to the state of the art

A Variety of different defects and dangerous situations can be found in rail vehicles occur, the various possible have negative consequences, by a faster wear of the track body and of the rail vehicle to serious accidents such as derailment, fire and the release of dangerous Rich materials. It is known, for example, that failure Of an axle box of a rail vehicle often to a derailment of the train concerned. Certain failures or insufficient functionality of braking systems can for overheating from one or more wheels and lead to their break, which often results in derailment of the train. It can be also overheat some brake parts themselves and, in some cases, cause a fire on the vehicle underside, causing the outbreak a very dangerous one large fire to lead can. Too strong braking force through the axle on the wheels can a shift of these wheels on the rail head and therefore a wear of the wheel profile and cause the formation of a flat, which in turn damages the track body and trigger a brake break can. Other wheel defects, e.g. A "welding on" can occur and cause a rapid wear of the track body. An excessive one Wear of the wheel profile can increase the rolling motion a bogie and therefore a rapid wear of the rail body and a possible Lead derailment along a long curve.

A excessive load on an axle or a bogie or a damaged entire wagons the track body and can in some cases lead to train derailment. An insufficient Securing the freight on a wagon or damaging one Fuse can cause that the loading cargo slips and falls on the adjacent track body or slides into a position that leads to a collision with another Train or with part of the infrastructure. That's out of a slipped Cargo resulting imbalance of a wagon can even one Slow overturning of the affected vehicle and the consequent Cause derailment. Unintentional opening of a wagon door or a Luke, an inappropriate loading of a freight car or the presence a combined means of transport with an illegal track image In a given section of the track, reasons for collisions may also be raised Share with other trains or with infrastructure elements. Unsecured parts of transport loads like that like a hood a car transported on a wagon can with the catenary in contact come and different possible damage cause. It can on board locomotives, wagons and railcars fire in a row Variety of random Causes or in the case of an arson that may break out size Losses on assets, Income and even human life.

In recent years there has been increasing interest in the railway sector in the use of technological systems which can detect the presence of certain defects and dangerous conditions for rolling stock. One of the main reasons for this, of course, is the interest in reducing the infrastructure and maintenance costs of rail vehicles, which still account for a large part of the total operating costs of rail transport systems. The significant reductions in many countries over the past few years in rail infrastructure and service center maintenance have generally made such detection systems more meaningful, at least because they have reduced the likelihood of defects and hazardous conditions by inspection or accidental personnel Observation to be discovered. In addition, the reduction of maintenance budgets for rails vehicles in some countries or in some railway companies has led to a more frequent occurrence of dangerous damage in the vehicles of the affected fleets.

Of course it is it possible or at least conceivable, many of the above defects and dangerous ones Conditions through technological systems that are based on the individual Vehicles are based to discover and most of the newer vehicles, especially the high-speed trains and locomotives in general, are in ever stronger Dimensions with a more diverse one and more effective selection of diagnostic systems. The costs for the retrofit of existing vehicles with diagnostic systems, however, are despite their considerable average lifespan, so high that a massive installation of on-board detection systems on not to the existing fleets, especially on freight vehicles recommend or exclude completely is. Accordingly, floor systems continue to be considered a cheap alternative and considerable research and development work has the improvement of existing products as well as the development of new products to the goal.

The Choice of positioning of ground equipment to detect defects and dangerous Conditions of rail vehicles is by no means coincidental, but is the result of a more or less careful decision-making process (for example, to the use of quantitative risk assessment methods), the on aspects like the performance of detection systems, the average Duration after which a detectable defect triggers an accident, the Probability that a certain defect leads to a major accident, the Meaning of a defect in the damage of track bodies, the frequency a defect, etc. is based. The cost of the application and maintenance of a recognition system are natural an important aspect and the installation of several series one certain ground-based detection equipment is often gradual for a Rail network, and for certain installation sites have a higher priority than others. A long railway tunnel or a series of tunnels along a railway line section a typical example of such a priority, those related to the higher ones average and maximum losses of the same accident (For example, a fire or a derailment or a derailment with followed by Fire), depending on whether this in a tunnel or along a normal open track section happens. The decision, certain Systems for the detection of defects or dangerous conditions of rolling stock near use of tunnels as a means of reducing risks [063, 064] (The numbers in square brackets refer to the numbered ones Documents in the reference list of sections 0, 0 or 0 continue at the bottom of the text), but should take into account the safety function, the such system for fulfill the entire railway network. Furthermore can more track sections with a more than average usefulness of these recognition systems, such as, for example Rail routes through heavily populated Areas nearby of shunting yards or ports or industrial areas where many wagons with dangerous goods start their journey. Long railway bridges are another example for one reinforced critical track section.

Of course, the sensitivity of a system to detect defects or hazardous conditions of rail vehicles (ie, the success rate of positive detection of defects or hazardous conditions) is a critical aspect that justifies the cost of installing and maintaining such a system. It is equally important to acknowledge that the frequency of false alarms is at least as critical as the sensitivity of the system to assess whether a system is actually to be used in the railway sector [064]. The detection systems mentioned here are indeed hazard reduction systems, in contrast vital elements of rail safety and signaling systems that must operate faultlessly. These hazard reduction systems are generally not mandatory, but when installed, their alarms indicating a dangerous fault or condition can not be ignored, and therefore such an alarm will trigger the stopping or rerouting of the train in question, too if such an alarm signal is "spurious", ie it does not indicate a real faulty or dangerous condition, it is therefore not surprising that the tolerance threshold for false alarms at railway companies is extremely low, given that the very high cost, incurred as a result of a false alarm, such as the additional costs of interrupting the scheduled train movement of many trains, delaying the train concerned, and wasting the time of the staff inspecting the stopped or diverted train technological systems Detecting bad or dangerous situations is a compromise between sensitivity and frequency of false alarms. Increased sensitivity for a particular technological detection solution generally results in a higher number of false alarms. Therefore, the maximum sensitivity of most detection systems to be installed on railway tracks results from a system setting for which the approximate maximum acceptable number of false alarms is crucial. These reflections are reflected in the fact that the most profound progress of past, present (and presumably future) recognitions Defective or hazardous conditions (ie, the complex development of the technology of "detection of hot axle sockets") are related to the improvement of the discrimination of systems between normal and abnormal objects [006, 007, 010, 011, 025], and all this with the Setting a low false alarm rate and sustainable costs It is well known to those skilled in the art that these improvements will be achieved, in part, through the use of better or different sensors, and in part through the use of better and / or alternative methods of processing the basic signals Sensors The topic of false alarms, which has been briefly addressed herein, is at the center of discussion about certain limitations of the prior art and some advantages of the invention discussed in this patent document.

In The following sections 0, 0 and 0 will be some specific previous ones Systems and documents explained, the most important for This invention is in the general category of those discussed above Recognition systems include. Further details regarding of the prior art, to which this invention is a cheap alternative represents, as well as the state of the art, in the context of the application This invention is or can be used are in the detailed Explanation in the Part 5 of this document.

2.2 Previous status of Technology for the detection of dangerous Deviations from the gauge line profile of rail vehicles

As mentioned above, there may be various causes (eg, an open door or hatch, slippage of a load on a flat railcar, inadequate loading of an open car, displacement of an unsecured cargo, and the presence of a wide load or a combined transport element associated with the Infrastructure profile of a particular section of track is incompatible) lead to the collision of a project from a vehicle too far projecting part with another train or with an element of the infrastructure. In the past, a number of accidents of this kind have occurred, despite railway companies' efforts to prevent them mainly through organizational measures and field tests by railway staff (eg at border crossings, at wagon loading stations and at marshalling yards). The automatic detection of dangerous deviations from the gauge profile of passing rail vehicles is therefore of interest to railway companies, but the development of an adequate technological system for this purpose presents extraordinary difficulties, which are due to a number of technically complex contexts within the leaflets UIC Code 505-1 , 505-4, 505-5 and 506 [050, 051, 052, 053], as well as other publications of a technical nature on this subject. The applicant considered it unnecessary to discuss in detail the content of these UIC standards in this document and the introduction of section 5.9, together with 15a and 15b , can simply be considered as a short reference on the subject, by those who already know it in detail. The discussion below should only be considered as a very brief and incomplete discussion of the subject, which confines itself to explaining some of the important reasons why the prior art known to the Applicant was unable to address the problem of detecting dangerous deviations from Track profile of passing rail vehicles, taking into account that the rate of false alarms must be kept at a low acceptable level for railway companies.

A first deciding factor for a possible collision of a vehicle body or an object mounted on the vehicle with an element of the infrastructure is the effect of the rail bend. In a typical two-piece wagon, a simple sketch shows (see eg 2 UIC journal 505-1 [050]) that the car box car on a curved track body assumes a position that depends on the positions of the bolts of the bogie, in which the car body is mounted. The portion between the bogie pins of the wagon side, which points to the inside of the rail curve, will point from the center line of the rails towards the center of curvature. Conversely, the parts on the opposite side of the wagon box, which are outside the section between bogie bolts to the outside of the wagon, in the opposite direction. The different geometric offsets of the various wagon parts in a curve leads to a "kinematic width" of a wagon profile on a plane which is perpendicular to the rails, which depends on the curve radius, the vehicle body profile and the position of the bogie bolts Therefore, still with respect to this simple case and assuming a certain amount, a simple, square cuboid shape with a certain lateral margin for the infrastructure of a curved rail guide will depend on its width and length as well as the distance of the bogie cast bolts from the outside of the vehicle body lateral travel and a rail bend, allows the shortening of the vehicle, this within the given infrastructure limits too broaden. In fact, longer rail vehicles are usually narrower than shorter ones. The rounding at the four vertical corners of such an ideal body makes it possible to increase its length at a given width, or to increase its width at a given length; this type of rounding is used in practice, for example, to increase the loading volume of freight vehicles within the approved gauge line profile (see 15b and Section 0, later in this document). The lateral displacement associated with a particular rail vehicle on a rail having a certain radius of curvature depends precisely on the longitudinal position on the vehicle, and therefore a load which accidentally or intentionally projects beyond the vehicle side (eg a flat body) ) yes, according to the position of the load on the loading deck, either collide with the infrastructure next to the rails, or not.

This last aspect will be along with other details at the time of authorization for the Transit is judged by wide loads, as it is applicable standard charging get hurt. The aspect explained here the kinematic width with respect to the rail curvature is used in the text below the explanation the restrictions, the solutions described below with the prior art for the detection of dangerous Mark deviations from the boundary line profile.

The Procedures used at loading stations and other train stations, such as marshalling yards, to measure vehicle profiles and their loads will not be here Discussed as they are for the automatic detection of dangerous deviations from the boundary line profile of passing vehicles are not crucial. From the same reason, patent documents such as [044, 045, 046] are not discussed here, since they are on exams, which are performed by hand or partially automatically relate, nonetheless with automatic testing the compatibility of gauge lines of passing vehicles not are compatible.

The Patent document [031] explained a system for detecting dangerous Deviations from the gauge line profile of rail vehicles, which registers the interruption of one or more detection beams (what the transfer of electromagnetic radiation or of acoustic waves along a Path through devices like mirror equals a ray in a series of straight Break jet segments), which are placed so that they one correspond to certain polygonal limit profile. A similar arrangement with a series of registration rays, each through a transmission and a receiver is implemented in the "CCD-1 Car Clearance Detection System "by General Electric Transportation Systems [963] used. According to document [066] based an electro-optical system for detection of dangerous Deviations from the boundary line profile of the company TSS from Milan, Italy, on the detection of protruding vehicle structures two lateral vertical planes and one horizontal plane the vehicle. Patent document [004] explains a system for detecting, in front of a railway tunnel, different risks for a railway vehicle, including the shift of the load, thanks to the interruption vertical laser barriers on the sides of the rail body in the Inside the "measuring tunnel", if provided, is recognized. More than one pair of laser barriers will be at different Positions along the railway tracks used with the aim of a To create redundancy and prevent false alarms, also by that the interruption times of the various barriers to the Train speed are tuned.

The four systems used in the o.g. Documents [031, 963, 066, 004] were mentioned, have in common that they are the disorder of the boundary line profile by the interruption of fixed detection beams detect, which are positioned along the rail track. If we assume that these instruments are installed on a straight railway track and that only vehicles with a standard load are the test point happen, and that the distance of the vertical detection beams from the middle of the rail is sufficiently large to trigger a false alarm for the largest width the for Transit vehicles registered on the appropriate railway track to prevent, then becomes in the case of such vertical detection beams at certain disturbances of the gauge line profile (e.g., a light but dangerous Load shift) on longer and narrower wagons occur, no alarm triggered. If however, this distance of the corresponding vertical detection beam shortened from the middle of the rail becomes, the number of false alarms is getting bigger. On top of that, often a considerable one Number of non-standard "wide loads" or "exceptional Loads "on rail vehicles transported and, as mentioned above, these loads on the Survive vehicle sides.

Consequently, a system based on detection by beam interruption with a given geometric profile could trigger false alarms during special transports. Thus, increasing the distance from the rail center to the vertical detection beam would be to prevent the triggering alarms due to abnormal cargoes cause a large number of dangerous charging situations not to be detected, if this distance is determined by the maximum width of the vehicles.

In In view of the objections described above, the four systems of the Documents [031, 963, 066, 004] are proposed to be dangerous Deviations from the boundary line profile to discover, provided they be installed in front of rail sections that are no or only very have flat curves (e.g., certain railway lines in large, wide Levels), and the sides of the polygons are correctly positioned for recognition. This limited applicability these four systems would but no solution with high detection rate and low rate of false alarms, because of the kinematic explained in the series 505 of the UIC leaflets Aspects of other vehicles. In particular, not even on one straight line the lateral displacement of vehicle parts in relation constant on the railway tracks, mainly because of the transverse axis play and the swinging vibration of the vehicle body. The recognition performance of dangerous Deviations from the boundary line profile of such systems is by the instability of lateral offsetting in the context of the default low rate Alarms restricted.

The Patent Document [047] a method and a device for measuring "distances of execution models" of car bodies Passing a gate on which a distance scanner is installed, and these execution models are compared with one or more predetermined execution models. His commitment is for a variety of possible Protective and / or safety applications recommended, such as the distinction of persons and vehicles at an open Entrance to a construction site. The installation of a suitable version of the device, which is explained in document [047] is at railway tracks would enable the determination of contour values of rail vehicles, which contain more information than the radiation interruption data, which are supplied by the systems described above, but that Document [047] does not provide information on a complete procedure, which the specific problem of detection of dangerous Deviations from the gauge line profile of rail vehicles with high discrimination performance (positive identification of dangerous Deviations from the gauge line profile) and a low rate release false alarms can.

The explained above restrictions The prior art is directly related with a precise distinction of impermissible vehicle profiles, according to the specifications the 505 series of UIC leaflets. However, the prior art is also insufficiently related with the identification of a loading shift during the test by the recognition system, which does not identify an invalid profile Has. Loading instructions, as well as the regulations, in the context of Application of the RIV agreement [060] apply, in fact, prescribe limits of width, which do not correspond to a fixed value, and instead of the Hang the geometry of the vehicle and also the one explained above Consider the lateral displacement aspect of a rail bend. Therefore, the prior art presented above is not suitable for a possible "light Shift "one Load to identify the consequence of violation of a loading instruction which concerns the lateral load profile.

Certain Loading rules govern other aspects (which are not the aspect of width discussed here), as well as the minimum distance between adjacent wagons when one of the loads over both Wagon box stretches. A violation of this rule may be a Load shift in length or a potentially dangerous one Display condition, especially with regard to the loading stability. A solution of this Problems are also not provided by the prior art.

2.3 Previous status of Technology for the detection of overheating in the axis area

The detection of abnormal heating or overheating of axle bearings has been the subject of intensive research and development efforts in the 1950's [005] in connection with the relatively high frequency of axle failures and the severity of their consequences (very often derailments). The failure mechanisms of bearings, and in particular the various types of towbar bearings, have been extensively studied (especially by bearing manufacturers) and are well known today. However, for the purpose of this text, it is sufficient to mention that axis failures are usually triggered by an exponentially increasing increase in friction caused by the simultaneous increase in the temperature of the bearing shell or other parts to which the heat generated by the bearings is directed, can be recognized. The incipient failure or the complete failure of a bearing can therefore be detected by the measurement of the heat radiation generated for example by a part of the axle bearing or by a suitable processing of these signals. Other methods of detecting axle failures have also been tested (eg by devices installed on the vehicle or by analysis of acoustic emissions from bearings), but none of these methods has until today been widely used as an alternative to measuring thermal radiation using devices installed along the railway line (often called "HBD", for Hot Box Detectors).

It many patent documents are over HBD developments (especially in the IPC class B61K9 / 06) and a detailed discussion this one goes far over the purpose of this text. In section 5.11 of this document Become some special HBD aspects, such as the species the infrared sensors, set out, and the text below Limited section insist on having some important features of the corresponding previous one State of the art to mention.

The first generation of HBD products was developed before modern Solid-state electronics is fed into everyday life, and the signals the sensor devices were sent to a leaf writer in a manned station, in which an engineer interpreted the received signals. Since became a considerable one Part of the developments on the automation of the analysis of thermal radiation signals directed in an attempt to improve the HBD sensitivity (in the sense that developing axle failure is detected as early as possible) in the frame to improve the required low rate of false alarms.

The Heat from Axle bearings are dissipated by conduction, convection and radiation. As e.g. in [49], is heat dissipation by convection a major factor in the determination of surface temperatures by a HBD registered heat radiation and therefore the temperature of such a surface is related to the ambient temperature a more appropriate single variable as the surface temperature alone for an early detection axle failure, as in some early HBD patent documents, such as in document [048].

The progressive use of ball bearings as a replacement for friction bearings was a challenge for the HBD industry, because the relative temperature (temperature in relation to the environment), at the a friction bearing is considered unfit for ball bearings represents the normal working condition. On top of that, ball bearings a larger permissible temperature range have, depending on the model and conditions of use. One of the first HBD patent documents [025] describes the use of wheel sensors, thanks to standardization from the wheelbase of freight cars, only for the scanning of bushings under Exclude exclusion of all signals from locomotives and passenger cars. The Patent Document [006] a method based on the processing of heat radiation signals of both Bearing the same axis is based. The same principle was in different Forms applied in other inventions to the variety of bearings and the effect of their recent "historical Sequence of operative conditions " to equalize their temperature. Patent Document [007] explains a system to distinguish between friction bearings and ball bearings based on their different shape, with the aim of hot bushings effective too detect.

The Patent Document [010] a method and system for avoiding the effect that the temperature a wheel hub (due to a normal or even faulty braking can be quite high) on the detection of overheating the bearing bush can have.

The Patent Document [011] describes a method for assigning wheels to rolling stock for the purpose of applying an adaptive threshold for HBD signals, by the average and the standard deviation of the signal values all bearings calculated on one side of the same rail vehicle become. Method of assigning wheels to rail vehicles were also explained by other patents to the processing of HBD sensor signals to improve [017], or to an alarm with the ordinal of a Axis and the ordinal number of the rail vehicle to associate, thus the manual test Axis failure is facilitated after a HBD alarm [023].

The Patent Document [008] goes up to the HBD detecting unit directed vertical measuring beam (instead of a series of obliquely directed Rays), and taken into account doing the fact that the rear in the running direction lateral surface the bearing bush usually cozy is as the corresponding front in the running direction, and that train service on a railway track usually in both Directions runs.

The Patent Document [014] the use of anamorphic optics to produce one Infrared-sensitive beam with an extending cut-out at the target or the use of an opto-mechanical scanner, both solutions Consider the transverse axis clearance.

The Patent Document [018] illustrates the use of a series of less pyroelectric ones Lithium tantalate sensor elements on a single chip with infrared image optics before, a plurality of heat radiation signals for many to produce different viewing angles. Some advantages of using Such an arrangement together with digital signal processing equipment explained here. Patent document [026] describes the use of a linear one Sensor die (especially a micro-thermopile die) with appropriate Image optics positioned laterally along the track is, with the sensor beams are mainly vertical and the sampling rate of the line is proportional to the wheel speed. Infrared images of passing wheels and bearings are detected and processed. There are many advantages compared to the previous one Prior art cited, including The possibility, both wheels and to accurately measure both bushing temperatures, compatibility with a Range of wheel diameters, the low energy consumption, the much lower exposure of the sensors to shocks and vibrations, precise temperature measurement, the permanent calibration stability, the insensitivity to fluctuations ambient temperature, easier alignment and insensitivity concerning the problem of the rear bearing seals of new bearings, the detection of certain commercial distributed HBD products by hot Bushings impossible do. The patent document [026] confirms the advantage of use of data processing techniques for the plurality of the linear matrix generates signals, but it does not provide detailed Details of the processing methods or algorithms, if applicable can be used opposite the prior art in terms of sensitivity in the context of the required low alarm rate an improvement to reach. Even if stock failure to the main incidents in the Related to axle-related components, so it is with wheels and Braking too safety-critical failure caused by a reasonable Detection of heat radiation and the processing of the corresponding signals can be detected. The Procedures and the system were accordingly developed to be individual Types of failure or dangerous Situations (for example, wheel overheating) or multiple types of failure or dangerous situations (e.g., detection a wheel overheating and a brake disk overheating through a single device) to recognize.

The Patent Document [012] describes a system for detecting the overheating of "each type of brake" by a single one Thermal radiation sensor, whose elevation and orientation angle are appropriately selected have to. Patent document [015] describes an apparatus of choice the orientation of the sensor beam, the detection of overheating both wheels an axis allows. Patent document [013] describes the application of a single one Thermal radiation detector, equipped with an optomechanical scanning unit that is not orthogonal to the railway tracks is installed and measuring the overheating of bearings, wheels and brake discs allows.

significant Progress has also been made in terms of disguise and the Installation of the sensors of the above detectors of defects Axial components made, especially in the attachment the sensor units (often called "scanner") on one Rail, the protection of the optics, the protection against snow and freezing and in the calibration and test equipment. For some of the latest equipment [963, 964], for example, the heat radiation sensors are combined with the most important optical and electronic parts in special accommodated deep sleepers and thus offer significant benefits in terms of robustness, ease of alignment, the protection against environmental influences and compatibility with maintenance work on the rails. Of the Prior art discussed above in this section of bearing failure and overheating of wheels and brake parts has indeed been a relative over the years Maturity attained. Further improvements are still possible before everything, but not exclusively, in the field of discrimination of abnormal and normal Conditions under the specification of a very low rate wrong Alarms.

2.4 Current status of Technology in the detection of fire on board rail vehicles

The first patent document [001] known to the Applicant with reference to FIGS automatic detection (by an apparatus operating on the infrastructure mounted) of a fire on a passing rail vehicle explained the use of sensors, and especially ionization probes, on a signal bridge mounted be in to "hidden" glowfire in recognizable with coal laden rail car, which would be stopped before a coal-landfill or reach an area where steam boilers are installed.

Part A of Patent Document [002] illustrates a system whose goal is to detect fumes or flames on heavy duty vehicles ("HGVs") before they reach a closed area, especially a tunnel. The fire is triggered by sensors, eg Infra Red sensors and "infrared-sensitive image converter" are detected, which are mounted on a signal bridge, which extends over the predetermined HGV path. The system includes other aspects such as, in particular, video cameras for monitoring entry into the detection area from a control room and means for diverting a HGV to a fire extinguisher platform or for approving onward travel on the intended route, depending on the outcome of the recognition process.

The Patent Document [003] explained a "thermographic System for controlling and preventing fire in a vehicle "that" a variety of Includes sensors that are supported by an arcuate structure and are suitable to the temperature of certain areas of a Vehicle. " Sensors are connected to a logical control unit that is "capable, at least to generate an alarm signal when passing through at least one sensor determined temperature exceeds the given values. "There are different solutions provided, namely the scanning of the vehicle by fixed sensors on a fixed Structure, through solid sensors on a moving structure and by moving sensors on a solid structure. In detail is provided that "visual Infrared sensors "are used and that the logical unit can trigger an alarm if it senses the sensor data compares with "thermal maps" of vehicles, registers the memory associated with the logical unit are. Document [065] describes an "infrared scanning system for the automatic Detection of overheating and starting fires in trucks approaching a tunnel. "There are two Versions of this system are explained and the second looks for scanning the vehicles before a series of apparatus, including one fast device for the Generation of infrared images and a fast device with one Linear CCD matrix for generating black and white images ("B / W") along with the associated Image management and processing units. The vehicle speed, the for the generation of images based on the results of the linear Imaging equipment need is measured by a special electro-optical apparatus. The process for generating the corresponding alarm includes a first step of classifying the "thermal properties" of the thermal images in a number of categories such as brakes, wheels, exhaust units, load volumes and upper cabin space. Special alarm criteria will be for the different ones Categories of thermal properties used and based on a statistical analysis of data from the recognition system itself. Document [066] also discusses the fire detection system described in document [065] and overheated objects in heavy goods vehicles on roads, before they enter a tunnel and provide more information the statistical performance of the system in terms of frequency of (real or false) alarms for different HGV classes. additional Information is about a similar System supplied that (for receiving and protecting the sensors) inside a railway tunnel has been developed and installed and the detection of potentially dangerous abnormal thermal conditions and incipient fires in passing Serves trains, the on a longer one Drive to the tunnel. The generating units of linear infrared images (IR) and VIS images ("visible") are in the network with Servers connected to the installation. There are a number of wheel sensors Installed along the railway track to approach a Train to detect a real-time estimate of train speed perform and to evaluate the relative positions of the axles in one go. The system classifies the elements with corresponding higher temperatures in categories, based on their morphology and position. If a fire or a very dangerous overheating of mechanical elements is detected, the system generates an alarm for the signaling and Safety system of the railway company, so that the train is stopped. A warning signal, however, is generated when a respect to his Heat not very critical abnormal element is discovered, so that at the next possible Stop a test can be made.

Patent document [004] describes a system for protection against accidents in tunnels. Sensors for at least one type of hazard are installed at a reasonable distance from the entrance to the railway tunnel so that a train with a defective rail car does not enter the tunnel but can be steered safely onto a protective stump. A "measurement tunnel" is installed at the level of the sensors in front of the normal railway tunnel and is designed to allow effective testing of gases, vapors or fumes from the passing train, and various methods of fire detection are described, including the use of one In addition, the installation of an infrared imaging device for each draw side intended to detect the "hot spots", so that "hidden" fires and / or smoke detectors are provided for train-derived air from the measurement tunnel and infrared and / or ultraviolet flame detectors The most suitable are infrared radiation sensors (at least in the context of the sensors mentioned for the above-mentioned fire detection) to detect (by apparatuses installed on the railway) an incipient fouling or flames inside railway wagons , as rising clouds of smoke are often not au to be detected by smoke analyzers, and because direct observation of flames or glowing surfaces is usually not possible. The detection of this type of fire is very important as it is often within a few minutes or within less than 10-minute intervals from the moment of sampling by a detector to a fully developed large fire, ie when the vehicle has already reached the very dangerous section of the track (eg the tunnel or marshalling yard in which many other wagons can be located with dangerous goods). Reliable detection of these fires by the processing of infrared delivery data requires, in order to prevent an unacceptable frequency of false alarms, a procedure that "hot spots" that belong to actual hidden fires, clearly from other points that are "warmer than the average "and are normally present in railway wagons (eg electrical appliances, heating systems, refrigeration, compressors of refrigerated wagons and heat exchangers, diesel engines and exhaust gases from locomotives, resistance regulators, kitchen ovens, restaurant wagons, etc.), as well as a variety of positions, Morphologies and intensity levels, given the wide variety of rail vehicles considered. Also, the identification of a hot spot due to a defect in an electrical circuit or the detection of an abnormally high temperature at a representative location of a locomotive surface requires the ability to have the relevant thermal characteristics of the variety of thermal properties unrelated to the occurrence of a hazardous situation to distinguish. The aim is therefore to overcome the above-mentioned limitations of the prior art in the detection of fires or overheating on board rail vehicles by new methods with an increased discrimination capacity between abnormal or normal elevated temperature characteristics, so that a higher detection sensitivity is achieved Rate of false alarms remains reasonably low.

3. Description of the invention

3.1 Object of the invention

The The main object of the invention is detection in one or more passing rail vehicles of one or more faulty and dangerous Conditions, specifically the dangers of the gauge line profile, suspended loads, overheating, losses and incipient failures in the axle bearings, overheating the wheels and brakes, overheating parts of the vehicle body and fire on board, by the application of the document described in this document and in claims 1 and 42 defined method (collectively "method") and system (collectively "system").

Moreover, by achieving the above main object, this invention also seeks to meet a number of requirements and / or benefits, in particular the following:
to generate a sufficiently low number of false alarms that is acceptable to railway companies, taking into account the risk reduction and the reduction in maintenance costs resulting from the installation and operation of the system;
accurately and with a high hit rate to detect the occurrence of impermissible vehicle and freight profiles in the context of the corresponding UIC code leaflets [050, 051, 052, 053] and so the limitations of the prior art, which are described in Section 0, to overcome;
identify clearly and efficiently geometrical features that do not comply with applicable charging rules and that may be indicative of a potential hazard, such as a possible cargo shift that (at the time of scanning the vehicle has not yet resulted in an inadmissible gauge line profile), or others abnormal cargo characteristics, thus overcoming the limitations of the prior art described in Section 0;
to detect the occurrence of fires aboard rail vehicles of all kinds and the presence of abnormal heat conditions in certain vehicles, eg locomotives, thus overcoming the limitations of the prior art described in Section 0;
To detect failures or incipient failures of axle bearings and the abnormal heating of wheels and of certain brake parts such as the brake discs and so to achieve an improvement over the prior art, described in Section 0;
an advantage, if desired, from the possible availability of a vehicle identification system that does not belong to the system (eg, by vehicle-mounted license plates or special marker panels or communicating with one or more vehicles in a train or satellite-controlled location or wireless Telephone technology of more or less complex logistics information systems), but which can be integrated into the system, but do not necessarily demand such availability;
having the ability to integrate the system with a variety of existing or future rail signaling and security systems or subsystems, which may require only minor further development or upgrades of some system implementation modules;
Assist crew members (train crews and train service crews) in their direct assessment of the occurrence of faulty or dangerous conditions by providing the system with accurate information on the nature of the identified hazard characteristics, individual vehicle information and position on the affected vehicle;
to enable the vast updating and management of maintenance software and permanent data to be performed without the need for any operator intervention in the system installations along the railways;
to be generally compatible with each vehicle type and model and adapt to new vehicle models without the need for system modifications, except for minor changes in permanent system data and possibly in some software modules;
to guarantee a high availability of system installations, thanks to the possible use of redundant and backup communication means to monitor such installations, including satellites and mobile telecommunications infrastructures;
to enable the direct and indirect integration of system installations with maintenance-specific information systems, and in particular with rail vehicle maintenance management systems, in order to reduce the average maintenance costs for rolling stock and the rail infrastructure;
perform the automatic detection of certain safety-critical defects and very dangerous conditions in passing rail vehicles, such as the detection of fire on board, inadmissible vehicle and cargo profiles, faulty axle bearings and overheated wheels and brakes, ie without intervention or Judgment by a person is required;
to provide the relevant information to a caregiver, such as an operator in a railway control center, to allow it to assess the occurrence of certain defects or hazardous conditions in a passing vehicle, such as a loose waggon sheet or a load shift occurring in within the allowable profile thresholds for which highly sensitive, fully automatic detection by a system implementation could trigger too many false alarms;
in the implementation of the system, to combine the appropriate hardware and software elements to discover, as described, a variety of defects and hazardous conditions, and thus the cost-benefit ratio is more advantageous compared to the installation and operation of different systems, because certain costs are shared and the information generated by the corresponding hardware and software elements are used synergistically;
to carry out the corresponding detection functions for vehicles passing the point at which the system sensors are installed at a speed of up to 120 km / h (maximum speed of most freight wagons and most freight trains throughout Europe) or, if desired, also for a higher driving speed;
Carry out the corresponding detection functions for vehicles, without the system sensors must be installed on a straight track, ie, if that is easier, they can also be installed on a curvature of the railway tracks;
to carry out the corresponding vehicle recognition functions, when the system sensors are installed on a straight line on which there are one or more adjacent rails whose distance from lane center to lane center is not greater than the distances normally found at main lines, ie at least 4000 mm most European lines with a standard gauge of 1435 mm [062], ie without a greater distance than normal between adjacent tracks;
perform the appropriate vehicle recognition functions by installing the system sensors so that they do not pose safety or operational issues despite their proximity to high-voltage overhead rails for single rails or multiple adjacent rails;
to carry out the corresponding detection functions for vehicles with a system sensor installation which avoids significant difficulties in carrying out ordinary and exceptional rail maintenance, in particular rail grinding, track daming and leveling of the tracks;
to carry out the corresponding detection functions for vehicles without the need to install any system sensor inside a railway tunnel or under a protective tunnel built for this purpose;
perform the corresponding vehicle recognition functions without having to carry any system sensor by a particular type of structure, such as a signal bridge or arched structure;
to carry out the corresponding detection functions for vehicles in any kind of natural light conditions and in any kind of weather, with the possible exception of exceptionally inhospitable weather conditions;
to perform the appropriate vehicle recognition functions without excluding any geographic orientation of the railway at which the system sensors are installed;
to enable the integration of the system with various apparatuses and systems whose sensors can be installed in the vicinity of the system sensors, such as different systems and apparatus for measurements and / or with other or redundant detection functions compared to those described above in the description of The main objectives of this invention listed detection functions, and especially apparatus or systems that measure the wheel or axle load or the load of entire rail vehicles and / or wheel defects and / or pantograph defects and / or other defects for which a suitable system exists, but not the required Recognize features for integration with the system;
Carry out the corresponding detection functions for vehicles in such a way that persons (passengers and / or train crews) are not exposed to infrared, invisible or ultraviolet radiation in accordance with the applicable safety standards and regulations without carrying out the relevant fault detection functions of these rolling stock with one or more persons is prevented on board or possibly on board;
the execution of the relevant detection functions of vehicles in a period in which the vehicles pass completely through the place where the system sensors are installed, which can thereby be shortened, albeit to a certain extent, that the system has a higher number units of account and / or faster or more efficient units of account, if necessary because of the proximity to a critical location along the track rail and to the location where the system sensors are installed.

One Another object of this invention is a method and system with a number of alternative options regarding certain Methods in the process as well as certain hardware and To deliver software elements within the system, which makes it possible to implement the system in one or more versions, it adapt to the individual rail conditions, different detection functions gradually implement and future improvements and / or the "fine tuning" of detection methods to facilitate.

3.2 General description the invention

This section of this document contains a general and introductory description of the invention and more particularly 1 and 2 , However, some important features of the method and system, including the subject matter of certain patent claims, are only explained in the following part 5 (Mode for Carrying out the Invention) of this description of the invention. Section 6 of this document contains a word dictionary with the meanings of acronyms, abbreviations and terms used together with numbered references to cited publications, standards and information about companies and products mentioned in this specification.

rail vehicles are usually associated with a vehicle execution model that the majority their characteristics like wheels, Axles, bearings, brakes, suspensions, Bogies, buffers, couplings, drive, body, ceiling, doors, windows, Hatches, electrical system, hot and cooling system Are defined. In contrast, the differences between parts a rail vehicle that is the same vehicle design model corresponds, very small, so for example color, overprints on the Pages such as vehicle owner's symbols or facility details of passenger carriages. Another difference between vehicles of the same vehicle model is natural the vehicle load, which is more or less clearly visible. New models of rail vehicles are undergoing a series of tests and acceptance tests exposed before being regular railways and the applicable regulations leave major changes on a removed rail vehicle not to. Accordingly are also subject to the interchangeable components of a vehicle Standards, which clearly affects the safety of rail vehicles and the procurement of spare parts and related logistics effect. In the text below, the term "vehicle identification" is used to denote a Process to describe the vehicle design model of a rail vehicle and possibly but not necessarily, even the unique identity of a recognizes certain part of a rail vehicle.

One The main feature of the procedure is that it is the one associated with the vehicle Construction model uses to diagnose defects and dangerous Defining conditions, it uses appropriate methods and parameters stored in a "vehicle database" and if necessary for the respective vehicle execution model can be retrieved.

The on the railway track based system for the detection of defects and dangerous Conditions in passing rail vehicles is based on the implementation, by immovable instruments and sensors, certain measurements of Parts of the passing rail vehicle and on the use of this Measurements as input for Software diagnostic applications. Another key feature of the procedure is the exact determination of the position and the orientation of certain Main components of a vehicle (especially of car bodies and wheel sets) Over time, so said measurements with certain parts of the Vehicle or its cargo can be associated, thanks to the by the the identification of the vehicle model is available Geometry of the vehicle possible is.

1 is a simplified diagram illustrating the process using information and data flow (boxes with rounded corners) to and from some processes or process groups (rectangular boxes). A train formation of vehicles (or even a single vehicle) 151 happens in one direction 152 the place where the sensors and measuring instruments 153 of the system are installed. The data 156 ("MEASUREMENT DATA") corresponding to the sensor signals and the instrument measurements are passed through the process 154 (Acceptance Measurements ") and stored in digital form for processing by the system's computational units The data transfer, as explained above in particular in Section 0, is done in such a way that all data is precisely, directly or indirectly, with a specific time value can be associated.

Once enough data has been taken, the process can 157 ("VEHICLE IDENTIFICATION MODEL") begin processing the inherited data to more and more accurately identify the vehicle design model of the vehicles passing the location on the railway tracks where the sensors and gauges are installed. Section 0 describes in detail a high success rate vehicle identification method that can be executed quickly, while section 0 describes another model that can be used for the few vehicles for which no execution model could be identified using the first method of identification. The methods described here for the identification of the execution models of all individual vehicles as well 155 train formation does not require special license plates or signs or markings on the vehicles, nor does the system receive any other information from external other systems outside the system, as elsewhere the application of the method would be conditioned by the availability of equipment and systems intended for Time can be used only for a small part of existing rail vehicles or existing railway lines. The vehicle identification method described in Section 0 uses as the main input the distances between wheelsets, the marking codes usually found on the vehicles (in particular the UIC marking codes of Bulletin 438 reference sheets [057, 058, 059]) and a set of data and information the vehicle models. Other inherited data may be used in the context of these vehicle identification methods described in Sections 0 and 0, such as vehicle weight, fast and accurate laser distance meter data as described in Section 0, and data from other electro-optical instruments.

The process on which the box 157 Also includes the determination of precise values for the distances between wheelsets ("WSD" for Wheelsets Distances) and the function ("LDF for Longitudinal Displacement Function"), which determines the longitudinal position with respect to a time of a vehicle along the railway line. The exact determination of the WSD and the availability of the LDF are required for the application of the identification methods described in Sections 0 and 0. In addition, the LDF is used for other processes in the process, in particular for the functions according to Box 160 ("VCPO FUNCTIONS PARAMETER COMPUTING"). A precise and safe method is explained in detail in Section 0 for WSD and LDF calculation based on the transit times of the wheels detected by wheel set sensors installed on the rails; This method is also compatible with the highest values for the positive and negative acceleration of passing vehicles. As explained in Section 0, WSD and LDF are preferably determined for the positions of wheelset centers instead of wheels, as these are slightly affected by the heavy fluctuations of bogies.

A set of vehicle-specific information and data 162 ("VEHICLE SPECIFIC INFORMATION AND DATA") are through the process 159 ("RETRIEVE VEHICLE SPECIFIC INFORMATION AND DATA") from the vehicle database 161 ("VEHICLES DATABASE") for each vehicle for which the vehicle identification process 157 a suitable execution model 158 ("VEHICLE CONSTRUCTION MODEL"). Of course, the vehicle database is one of the most important elements of the system. It will be discussed as an issue in Section 0, while the individual content will be addressed in various places in Part 5 in relation to the variety of different methods used in the process. The box 160 refers to the calculation of the parameters 163 ("VCPO FUNCTIONS PARAME TER ") that define certain mathematical functions and calculate the position and orientation in a ground-based coordinate system from a main constituent of a vehicle whose execution model has been identified (VCPO stands for Vehicle Constituent Position and Orientation) These mathematical functions correspond to time-dependent coordinates Conversion functions between a ground-based coordinate system complete with sensors and measuring instruments and a coordinate system complete with a main component of a vehicle Such coordinate conversions are a key element for the implementation of various functions described in Part 5 for detecting defects and dangerous conditions in a vehicle Model, since they allow the relationship between an acquired measurement and an element of a main component of a vehicle to be established, depending on the nature of the measurements s (eg, value from a scanning laser rangefinder or from an imaging device), the reference is defined directly by a vector in three dimensions or by the combination of a three-dimensional vector with a target surface. Section 0 describes the details of a method for calculating the VCPO function of the van of a rail vehicle whose model has been identified, using the matrix characters for rotation (RPY, Roll, Pitch & Yaw, rotation angle of the camera around sight axis in degrees, inclination angle in Degree, azimuth of the viewing direction in degrees) and their transmission into uniform coordinates. A number of sub-methods are presented to use various inherited data in combination with data and information from a vehicle database to estimate a set of mathematical expressions that can be used in the calculation of the relevant VCPO function using a multiparameter minimization algorithm is used. Similarly, the methods and sub-methods described in detail in Section 0 are specific to the calculation of a mathematical VCPO function for each wheel unit and the corresponding axle-related parts of a vehicle whose model has been identified.

The box 164 ("DEFECTS & HAZARDS DETECTION") refers to a collection of methods and processes that serve to detect specific faulty and dangerous conditions of a vehicle whose model has been identified and the inherited data associated with the vehicle, in general, to the vehicle VSPO functions and a specific set of data and information from the vehicle database. Section 0 describes a set of methods for detecting hazards associated with the boundary line profile of the box and the cargo of a vehicle whose model has been identified. The use of the VCPO function for the car body allows the precise determination of the position of three-dimensional points, which were determined by means of suitable instruments and algorithms on the car body. Thus, based on the vehicle-specific information and the data from the vehicle database, it is possible to determine whether certain three-dimensional characteristics of a vehicle and its cargo are inadmissible for the section of the train to which the train is approaching, the basic information given in the corresponding UIC Code Leaflets [050, 051, 052, 053]. The method may also take into account the specific infrastructure profile characteristics of a particular rail line section, as long as they are known and represent a constraint or tolerance compared to the standard infrastructure profile. Detection of hazards associated with the gauge line profile of combined transports (semitrailers on bogies, mobile bodies, HGV transport on wagons, etc.) is also possible as specified in the relevant UIC Code Leaflets [054, 055, 056]. Exemptions from the Freight Line Profile may also be discovered on the basis of the applicable codes and regulations, in particular the provisions of the RIV Agreement [060]. In the particular case of exceptional loads, the method allows for automatic recognition of its compatibility with the corresponding infrastructure gauge line profile, taking into account (if available) the train speeding schedule. The recognition and registration by OCR of special markings of combined transports and of coded special shipments makes it possible to detect special violations of the corresponding loading profiles. The use of data and information transmissions between a system installation and one or more lane information systems is described in Section 0 in connection with the detection of shifted loads, loose wagon panels and the speed dependency of loading line profiles of special transports. The detection of dangerous deviations from the boundary line profile on the underside of rail vehicles is explained separately in Section 0. Section 0 describes the details of methods for detecting defects and hazardous conditions of axle-related parts (especially bearings, wheels and brakes) of a vehicle whose model has been detected, by processing data from sensors and instruments corresponding to those of the vehicle Surface generated heat radiation register. Depending on which sensors and instruments are used in the system and for the specific FCPO function, and together with the vehicle-specific data and information from the vehicle database, different algorithms and alarm criteria allow more accurate differentiation of normal conditions from abnormal conditions as was possible with the prior art, and at a very low rate of false alarms. In particular, the recognition of the vehicle model, and thus the corresponding axle-related parts, enables optimum application of the alarm criteria, based on a statistical comparison of thermal data corresponding to the identical parts of the same vehicle and / or vehicle in the same train formation. The use of the specific VCPO function allows precise association of thermal emission measurements with parts such as the brake discs, since their existence, geometry and position are well known, as well as the existence, geometry and position of other parts in the foreground or background, so as registered by the corresponding instruments or sensors. It also discusses the possibility of improving the distinction between faulty and regular bearings by taking into account the corresponding mechanical loads that are based on weight measurements by a suitable system integrated into the system.

Of the Section 0 describes the details of detection methods of fire and abnormal heating conditions of the box and the Freight on vehicles whose execution model has been identified. The special VCPO function is used in this case to calculate heat output data to associate with elements of a vehicle body or its cargo, based on the vehicle-specific information and data the vehicle database. These methods are characterized by a series of sub-methods for the preparatory processing of the measurement data and algorithm data, which the Serve the evaluation of the results of preparatory data processing, and specifying these methods and algorithms, along with the corresponding parameters for a particular vehicle model are to be applied retrieved from the vehicle database. The featured in section 0 Methods apply to the detection of a wide variety of abnormal ones Heating conditions related to fire. In addition will be some representative Situations of fire dynamics and shown for specific vehicle types. The possibility the individual adaptation of the recognition process to different ones Vehicle types in a number of different parameters based Detection methods together with the precise association of measurements with elements of the vehicle and its cargo make it possible a significant improvement in detection sensitivity compared to achieve the state of the art, and this at a very low rate of false alarms. The described data processing methods are also suitable for the configuration of the detection process from other abnormal ones Heating situations (at the time of detection not one Match fire), especially on parts of the locomotive.

Of the Section 0 describes some possible ones System functions which specifically concern the railway transport of dangerous goods. The standard posters for dangerous goods, identify rail vehicles and combined means of transport, and the codes of the corresponding goods are included in the Recordings of the vehicle sides recognized and the association with the Weighing wheelsets and allow the vehicle weight from the vehicle database it to create a list of the vehicles concerned with the indication whether these are loaded or almost empty. This list, possibly some redundant information compared to other rail safety systems can be shipped or made available on request by other system or she may be related to the temporary or permanent Ban on the circulation of dangerous goods along certain sections be used in the railway network (for example, double-track tunnels through which trains drive).

weight and / or cargo measurements for Bikes, wheelsets and / or vehicles taken over by special apparatuses which could be integrated into the system, as explained in section 0. This data can especially from the system for it used, certain own processes for the detection of defects or dangerous Improving conditions (for example, distinguishing defective axleboxes) and / or to create specific damage diagnoses based on weight, by the weight values with the vehicle specific information be compared (e.g., to detect a particular disregard the maximum load per axle or per vehicle or imbalance in the load distribution on a vehicle).

Of the Section 0 explains in short, the possible Integration of additional sensors, systems and subsystems and the Advantage of sharing certain system features or system-related infrastructures and use of vehicle-specific Information from the system to improve the detection methods compared to the prior art (e.g., pantograph diagnostics) and / or the development of further innovative damage detection methods.

For the majority of vehicles, the identification process according to Box 157 also recognize the distinctive identity of a vehicle based on the marking codes, which, as described in Section 0 which can be used to integrate the system with maintenance management systems of rail vehicles and logistic information systems. The detection of defects and hazardous conditions in new vehicles whose model can not be identified is made using less accurate and / or reliable methods (Section 0) than the detection methods using vehicle-specific information and data from the vehicle database. 2 represents a simplified sketch of a typical system installation where a train formation 204 of rail vehicles on the tracks 202 and 203 the railway line 201 in the direction 209 to a line segment where sensors and instruments are installed. The dashed area 205 denotes the "SMI" (used here for system measurement interval "system mesasurement interval"), ie position interval along the railway tracks, at which a part of a passing rail vehicle can be measured by one or more sensors or instruments. The actual length 211 The SMI depends on a number of different factors, most notably which and how many instruments and sensors are installed. Section 0 details the aspect of SMI length in the installation area of wheel sensors and other sensors or instruments. The dashed areas 206 and 207 correspond to the position of the sensors for detecting the approach of a new train from one of the two possible directions as well as for calculating the approximate speed and approximate time at which the train reaches the SMI, allowing the system to read the data of the sensors positioned at the SMI and to take over instruments. Some system configurations, in principle, may not need any sensors in the "train detection areas" 206 and or 207 , The distances 210 and 212 If necessary, they must be sufficiently large enough to allow enough time for the preparation of certain apparatus (eg optical instruments with protective covers that can be opened or with rotating parts that do not move when the apparatus is stationary), so that they have their respective functions for the intended purpose execute approaching train. The connections 214 . 213 and 216 generally characterize the groups of connection means for the operation of the sensors and instruments in the areas 206 . 205 and 07 , and the box 208 represents a set of system apparatuses including data transfer units, data processing units, communication units, power units, etc. 215 identifies one or more cabinets or guards or shelters for the apparatus of the box 208 while the line 217 the power supply indicating signal and communication connections, with particular reference to the connection to the rail signaling and security systems and to one or more communication means with other systems and with centralized system remote operation sources, as described in detail in section 0. Those skilled in the relevant arts, given the detailed descriptions and explanations below, will recognize that the data transfer, data processing, communications and power equipment are very conveniently located and / or deployed, and in a more complex manner than in 2 can be connected (eg by installing some data transfer devices close to the corresponding sensors and instruments).

It can various combinations of sensors and instrument types for implementation of the method chosen and it can be a variable number of sensors and instruments for each of these Types for the takeover used by data entered into different processes become. The characteristics and advantages and disadvantages of certain sensor or instrument types are in different sections of part 5 explains as well as the respective installation options and the relationships between the use of certain sensors and the application one or more relevant methods under the procedure. Section 0 describes different wheel sensor types and the section 0 explained some different measurement uncertainties or errors that occur their use in the process. The section 0 goes especially on a family of fast and accurate laser distance sensors one in the procedure for the takeover of profile data of wheels and other parts on the vehicle underside of rail vehicles can be used. Section 0 describes various types of VIS and NIR imaging devices and especially of scan image generation devices for lines used in the process for the Recognition of vehicle marking codes and for other purposes such as determination the VCPO function, the recognition of dangerous Arrangements of goods and the supply of vehicle images to railway control centers can be used. Three - dimensional measurements of the positions of vehicle parts can be made in the Procedures are carried out in the framework of various types of instruments which are mentioned and described in section 0. In the process can different families of other instruments are used to heat outputs of vehicle parts and they are in section 0 about axis-related Components and in section 0 above car bodies described.

Section 0 describes some options for implementing the data acquisition of various types of sensors and instruments as part of the requirement of the procedure, one exactly time value directly or indirectly associated with each measurement. The calibration of sensors and instruments and of certain geometric features of the measuring instruments is discussed in Section 0. The subject of calibration is also dealt with in some other sections on the sensors and instruments and on the use of the data acquired by them through various methods and algorithms. The accommodation and installation aspects associated with sensors, instruments and electronic apparatus of the method are discussed as an issue in section 0., while section 0 describes some of the more preferable choices for the collection of software applications in the context of system implementation in particular, preparations for reducing as much as possible the time needed to complete the processing of a train, and for using high-efficiency computerized equipment.

Of the For simplicity, the text goes in the descriptions and the accompanying drawings in part 5 predominantly assuming that the railway line at the SMI is straight. Section 0 describes some less important aspects involved in system implementation considered Need to become, if the railway line at the SMI has a large curvature.

section 0 explained some examples of system configurations, with a special note on different combinations of sensors and instruments that be installed on the SMI.

3.3 Advantages of the invention

One first advantage of this invention over the prior art of Technique is that it provides a method and a system that is able to automatically generate an alarm when a vehicle and / or his freight for conditions for the boundary line profiles that apply to a particular line segment do not fulfill, the method being the kinematics of rail vehicles and before all of the car bodies considered.

In front can do anything the principles and details of the corresponding UIC Code Leaflets of Method and system used to alarm for a gauge line profile (as described in UIC codes 505-1, 505-4, 505-5 and 506 [050, 051, 052, 053] is defined) and also, if desired and applicable, the actual Obstacle profile of the track section and the intended speed the train in question. The details of the corresponding UIC code leaflets for combined rail transports (Leaflets 596-5, 596-6 and 597 [054, 055, 056]) can also be used in the process and included in the system become. The very low rate of false alarms typical of such Function of detection of impermissible Vehicle and freight profiles is possible a link system installation with the rail signaling and security system for one Significant reduction of risk of collision with vehicles of others Train formations on an adjacent track or with elements the railway infrastructure, and this risk reduction is achieved without the operationality in rail transport disturbed which, on the other hand, is increased Rate of false security alarms would be the case.

One second advantage of this invention over the prior art of Technique is that it provides a method and a system that is able to provide an alarm or warning (mostly dependent on the different possible ones Integrations with other rail safety systems and control centers) generate if a violation of loading rules and regulations present and / or in the case of a freight shift on a wagon, So if abnormal loading characteristics at the time of detection are not serious Danger in connection with the admissibility of boundary line profiles represent. Also in this case are alarms with a high discriminating ability generated between normal and abnormal conditions, so that the Quote false alarms is correspondingly low, as alarms instead of alarms are generated and also more information about possible faulty and / or dangerous Conditions (e.g., a loose waggon sheet), which is a Assessment by staff require.

One third advantage of this invention is that it is a method and deliver a system that is capable of the skills the prior art in the detection of faulty and dangerous Conditions of axle bearings, wheels and to further improve braking by reducing the heat dissipation of measures these parts.

A fourth advantage of this invention is that it provides a method and system capable of measuring heat radiation generated by the box and / or interior and / or cargo of a passing rail vehicle. the presence of a fire Detect board on the vehicle, and this usually with a much greater sensitivity compared to the prior art and especially in a fire in its initial stage, which is detected in a closed vehicle compartment; This detection of the presence of fire is on condition of a very low rate of false alarms.

One fifth Advantage of this invention is that they are a method and provides a system that is capable based on measurements the heat radiation, which is generated by a passing rail vehicle, and under the requirement of a very low rate of false alarms that Presence of a abnormal heating of vehicle parts, especially in locomotives, recognizing by applying data processing methods, which can be customized to specific vehicle specific faulty or dangerous Conditions with a much larger discrimination capacity compared to recognize the prior art.

One sixth advantage of this invention is the possibility Weight / load measurements of wheels and / or wheelsets to integrate and use these measurements in comparison to the prior art in a better way those vehicles to distinguish too large a weight on the wheels, axles, Have bogies and / or on the entire vehicle and / or have an uneven distribution of weight.

One seventh advantage of the method and the system is the possibility automatically and independently Create lists of vehicles that are within a train formation Transport dangerous goods, and possibly also indicate which transported goods lie on the respective vehicles and whether these respective vehicles as train tanker for Loaded with chemicals or almost empty, because such lists are used different security purposes and are in different integration schemes used to in accidents (e.g., derailment of train formation in a railway tunnel) are very important Retrieve information and / or train with dangerous goods thereof to deter, to reach a certain section of track, if this not permitted is (e.g., access to a twin-track tunnel, if the same Time a passenger car drives through the same tunnel).

One eighth advantage of the method and the system is the possibility the unmistakable identity of most rail vehicles in a train formation, without for this purpose, identification devices mounted on the vehicle or other non-procedural identification systems use and this distinctive identification information, with or without additional Information about detected Defective, more useful for delivery Information and data to maintenance management systems and supply logistics systems. A ninth advantage of the process and the system is the ability to the most suitable places for the installations in one To choose route network, and that independently of restrictions like the presence adjacent railway tracks, rail bends or the geographical Orientation of the rails. Professionals will get more benefits from this Procedure and this system, in the text of Part 5 below described in the document, appreciate.

4. Short description the drawings

1 FIG. 12 illustrates the basic elements of the method with reference to claim 1. Rectangular boxes with solid border lines correspond to the main processes, while boxes with solid border lines but rounded corners correspond to data and / or information.

2 is a very simplified sketch of a generic system installation.

3 shows a possible collection of the main software applications (rectangular boxes) implemented in the system. In general, such applications run asynchronously to each other and each receives and produces data and information.

4 is a simplified flow chart of tasks and software applications directly related to the acquisition of sensor and instrument measurement data.

5 FIG. 12 is a composite diagram used in the description of the method for illustrating the relationships between transit times determined by wheel sensors for the wheels of a vehicle, the distances between these wheels, and the longitudinal position of that vehicle with respect to time along the track rail.

6 shows the pendulum motion of a wheel in the bogie of a rail vehicle and relates to the effect of the yaw variation of bogies and their consequences on the transit time of individual wheels of the bogie and wheelset centers.

7a to 7h are used in the descriptive text to express the uncertainty of interpolation and extrapolation concerning the calculated position of vehicle elements with respect to time (according to a method described in this document), in the context of the distances and the range of wheel sensors, their signals used to calculate a function that expresses the vehicle position in relation to time.

8a and 8b show the positioning of a fast laser knife that can be used within the system to take over wheel profiles. 8c shows two examples of such profiles.

9 shows a simplified flow chart of a method described in this document for detecting the execution model of individual vehicles in a passing train formation of rail vehicles.

10a and 10b show possible approximate positions and orientations of VIS / NIR cameras used in the system to take over line images of the passing rail vehicle.

11a and 11b show possible approximate positions and orientations of solid laser distance measuring devices at the time of passing that can be used in the system to take three-dimensional measurements of the position of features in a passing railcar and its cargo.

12a and 12b show possible approximate locations and orientations of opto-mechanical laser scanning "time-of-flight" distance meters that can be used in the system to take three-dimensional measurements of the position of features in a passing rail vehicle and its cargo.

13a and 13b Figure 12 shows a possible approximate position and orientation of a high performance, high speed scanning distance measuring device that can be used in the system to take three-dimensional measurements of the position of features in a passing railcar and its cargo.

14 shows two three-dimensional coordinate systems, namely integral with the car body and the sensors and instruments of the system installation along the way with respect to the definition of a coordinate transformation of a vector / point between two such coordinate systems.

15a and 15b are each out 5 and 4 UIC Code Leaflet 505-1 [050] adapts and shows a series of profiles defined in this leaflet and UIC Code Leaflets 505-4 [051] and 505-5 [052], which describe the relationships between a reference line profile, the maximum permissible profile of a rail vehicle and the obstacle profile of a railway infrastructure.

16 Figure 3 shows two three-dimensional coordinate systems, each integral with the vehicle body and sensors and instruments of system installation along the way, relating to the comparison of the position of a three-dimensional position measurement of a vehicle part of its cargo with a tread set defined in the coordinate fields integral with the car body.

17a and 17b Figure 2 shows two views of an axle of a rail vehicle, including the wheels and brake disks, together with a linear infrared imaging device used to take heat release measurements for such mechanical components.

18a and 18b Figure 12 shows two views of an axle of a rail vehicle, including the wheels and brake disks, along with two three-dimensional coordinate systems, integral with the axis centerline and with the sensors and instruments of system installation along the way with respect to the definition of the coordinate transformation for a vector / point between two such coordinate systems.

19a and 19b are two simplified representations of the areas of intersection between measuring beams of an instrument for measuring the infrared emissions from surfaces, the outlines of the well-known position of such intersection areas, and three areas in the foreground, in the background, and at an immediate position, and these two drawings refer to Figs different ratios between the inclination of the measuring beam and the width of the known contours along the direction of a series of overlapping regions of measuring beams.

20a similar 19b but has no known outlines while 20b a diagram shows that the temperature measurements at the measuring points of 20a together with the actual temperature profile corresponding to the three corresponding surfaces along the direction of the measuring point centers with a curve that interpolates these temperature measurements.

21a and 21b show a possible indicative positioning and orientation of a linear infrared imaging instrument that may be used in the system to take heat release measurements of box parts and visible cargo on a passing rail vehicle, along with other instruments.

22 shows some possible connection schemes for the data transfer of the system by master data transfer units or other data transfer units and such connection schemes, they are suitable for direct or indirect, assign existing time values of each measurement, indeed characteristics of the various sensors and instruments.

23 shows some basic elements of a system installation and a number of remote elements, such as remote add-on components of the system or parts of other systems, as well as the connections between these elements.

24 is a bird's eye view of the sensors and instruments installed on the SMI for a sample configuration of a system installation.

5. Implementation possibilities the invention

5.1 Some general Features of the process and the system

The method involves a series of processes that perform various tasks such as measurements, data management, data processing, communication and signaling to adequately detect and signal a number of potentially erroneous and hazardous conditions. 3 is a general conceptual diagram that provides an overview of all of the said tasks that can be grouped into a realm of different single or compound processes (rectangular boxes numbered from 231 to 246 ), usually as input data / information that is part of a parent record ("DATA") 230 are, receive and produce as output. The output of a given process is usually the input of one or more other processes. Some of these data and / or information have a control function for certain processes because they define their tasks and / or trigger their activation.

Below is a brief commentary on each of the elements of 3 to provide an overview of the main part of the main system process; later, they will be further explained in the description of the method and system in other sections of this document. The applicant would like to make clear before the short comments below that the set of rectangular boxes in 3 is not necessarily exhaustive, nor does it correspond to the single or best set of processes to define the implementation of the method and system. In addition, the representation of the 3 based solely on the data and / or information flow, equivalent to other types of representation (eg, SADT [074]) of system processes in which controls and data flows can be shown differently, and especially if all or some of them are directly linked to process symbol pairs in association with.

box 232 ("TDA &RTDP", for Trigger Data Acquisition & Real Time Data Processing) corresponds to a software process similar to the ones in the location areas 206 and 207 in 2 and monitors some of the tasks involved in providing data and triggering cues or signals that prepare the system to "scan" a train with the instruments positioned on the SMI in real time, those purchased at the SMI processing data (eg changing the rate of data transfer, detecting insufficient train speed, etc.).

box 233 ("SMI data acquisition") refers to the process of data transfer of sensors and measuring instruments located at the position range 205 of the 2 are installed. The timing of this process after recognizing a move through the process of Box 323 is explained later in this section, while information on the instruments and sensors of the 4 in different following chapters of the document, such as 0, 0, 0, 0, 0, 0, 0, 0, 0, and 0. The features of the data transfer equipment are discussed in Section 5.18 of this document. It may be an advantage that all the data of the wheel sensors installed at the SMI is through the process of the box 232 rather than through the process of the box 233 to be recorded.

box 234 ("Primary VI process") describes a basic process and system process used to quickly and safely identify the execution model of a large portion of the vehicles that have passed the SMI and possibly unambiguously identify them. This process involves calculating the longitudinal position of vehicles along the railway tracks in relation to the time and the distances between wheelsets of the passing vehicles. The details are described in sections 0 and 0.

box 235 ("secondary VI process") describes a process that attempts to identify the execution model of those vehicles that are affected by the primary vehicle identification process of Box 234 could not be identified. The details are described in sections 0 and 0.

box 236 ("VCPO computing") describes data processing applications for calculating the functions that define the position and orientation of certain major components of vehicles in relation to time in a coordinate system integral with the track or railroad tracks. The VCPO calculation of those vehicles whose execution model has been identified is addressed in relation to car bodies in Section 0, while the VCPO calculation for axles and related components, also for identified vehicles, is explained in Section 0. box 237 ("VB gauge diagnostics") refers to the process of detecting erroneous or dangerous boundary line profile conditions of vehicles whose execution model has been identified. This process is described in Section 0.

box 238 ("VB thern.diagnosis") refers to the process applied to a vehicle whose execution model has been identified to detect faulty and dangerous conditions of a vehicle body, particularly fire and incipient fire on board, by analyzing the measurements of heat outputs. The process is described in Section 0.

box 239 ("BWB therm. Diagnostics") refers to the process applied to a vehicle whose execution model has been identified to detect overheating of faulty and hazardous conditions of axle-related components by analyzing the measurements of heat emissions, particularly with respect to incipient axle failures brakes and brake damage. The process is described in Section 0.

box 240 ("UV diagnostics") refers to the process of detecting a number of faulty and dangerous conditions in such vehicles whose execution model has not been identified. This process, including the corresponding VCPO calculation functions, is described in Section 0. box 241 ("RSS interfacing") refers to the process associated with coupling the system to other rail signaling and security systems to signal alarms and / or to exchange data and messages with railway information systems and control centers. The topics of alarm triggering as well as information that can be exchanged between the procedure and one or more systems or centers of the railway company are explained in various sections, eg 0, of this document, while the actual coupling solutions are described in section 0.

box 242 ("data transact.") refers to the process associated with a plurality of record transactions, particularly with respect to the sending of measurement records, record files, prepared data, and other records from a system installation to a remote data processing system for one or more purposes as well as the offline calculation for improvement one or more diagnostic methods used in the method implementation. The remote control of software and vehicle database update can also be understood as one of these processes. The processes of Box 242 are addressed in various parts of this document, especially in Section 0.

box 243 ("config. &calibration") refers to a set of functions related to the configuration of the process installation, from the appropriate calibration of the instruments and their geometric positioning and orientation to verifying the accuracy of the calibrations and any adjustments required. The functions of Box 243 are addressed in various parts of this document, especially in Section 0.

box 244 ("Reporting") refers to the processes related to completing and submitting reports in electronic form, ie sending diagnostic reports to a railway center or to a service crew, thus checking the detected erroneous and dangerous conditions and take appropriate corrective action. The processes of Box 244 are addressed in various parts of this document, especially in Section 0.

box 245 ("integration functions") refers to a set of optional processes for integration in the process of one or more systems, such as a wheel weighing system and a wheel detection system, and for managing such data from these systems for one or more purposes such as for example, the processing of diagnostic data and information by the fundamental process processes and said systems. These processes are addressed in several parts of this document (eg in section 0 on the use of wheel load data within diagnostics of incipient journal bearing failure), and more specifically in section 0.

box 246 ("HAZMAT Functions") refers to the process associated with the processing functions of records that may serve to reduce the hazards of dangerous goods on rails, as specifically described in Figure 0 of this document.

box 231 ("System supervisor") refers to the process or processes that can be implemented within the software of a process implementation to monitor and control process operation. The corresponding functions are briefly addressed in section 0 of this document.

box 230 refers to a plurality of data and information, including, and in particular, the data and information from the vehicle database, data from measurements made by the method, configuration and calibration data and parameters, results of data processing functions, alarm flags, parameters of the data VCPO functions, train train list, diagnostics reports, and other data and information used or produced by one or more data processing modules in the method implementation.

Those skilled in the art will recognize that some of the features of the method proposed by the Applicant in the following description concerning the structuring of the data processing are such as the relationship between total time for processing data for vehicle formation and cost improve the computing means of a system implementation. Most of the software processes that require considerable computing resources (eg those in the boxes 236 . 237 . 238 . 239 . 240 and 246 addressed) may be executed asynchronously once the necessary input data for a vehicle is available, and more than one example (ie, different examples of the same processing data of an application for different vehicles) may run in parallel on the same data processing unit or on different data processing units Overall, a considerable advantage for the use of the calculation means of the method means. Another function that can be used to optimize the use of the method's calculation means and speed up the delivery of the method's diagnostic responses is to trigger certain data processing tasks, such as boxing 234 . 235 and 236 which generate vital data for computationally intensive applications as soon as sufficient data is available, ie, without waiting for the completion of the acquisition of measurement data for an entire vehicle formation (for example, when asynchronous data processing applications do not start until measurements for a 500 m long train, which is scanned at an approximate speed of 50 km / h, are completed, then a delay of about 36 seconds at the time for completing the application of the diagnostic functions of the method is added). 4 describes a method used in arranging some of the train transit time It is preferable that bounded process tasks be such that the sampling of a train is not due to the completion of the data processing of an earlier sampled train and thus reduces the minimum time between sampling successive trains and / or possibly reducing equipment costs for data processing.

box 219 ("waiting for new train") corresponds to a "stand-by status" of the procedural instruments as well as the data transfer equipment and the software. In this status, no measurement is performed by the sensors and instruments installed on the SMI, except those designed to control the system's functionality, to validate current calibrations, and to perform other possible diagnostic processes (eg, detecting impairment of certain calibrators) optical sensors in very dense snowfall). A transition from box 219 to box 220 ("prepare to measure") occurs when a new train is detected, which is to be scanned, ideally by the wheel sensors connected to the in 2 With 206 and 207 marked positions are positioned. A function according to box 220 is the activation of the electric motors for those measuring instruments such as certain laser distance measuring devices, which have a rotating optics that does not turn when the instrument is stationary for a certain period of time. Another feature with box 220 could be opening the protective covers or flaps of certain optical instruments that have such protections to prevent dirt, dust, water or snow from settling on the optics when the instrument is stationary. There may be more actions with the box 220 depending on the instruments installed during the process installation. The time required to execute the actions started for the box 220 , along with the maximum speed assumed for a train approaching the SMI, may be the deciding factor for the minimum value of the distances 210 and or 212 from 2 be.

box 221 ("start new train job") is activated as soon as the corresponding commands in box 220 were entered. The with box 221 Associated tasks correspond to a series of initiations and settings of control indicators and variables for sampling and subsequent data processing for a new train. Of course, these tasks are defined in the detailed software design for a process implementation.

After completion of the tasks according to the box 221 (typically requiring a few tenths of a millisecond) a wait cycle begins on the box 222 ("wait train motion data"). This cycle ends with the transition to Box 223 if the information about the approach speed of the train to be scanned is available.

box 223 ("select scan parameters") refers to the choice of certain scanning parameters, which are defined according to the approach speed of the train and possibly also with reference to other information and data that can be measured or taken over by external information system, before the train the SMI Area reached.

One simple example of such a parameter setting could be that Omitting a data collection for the detection will be a flat of the wheel, if the train is too slow or omitting certain measurements (e.g., data from one slow laser scanner), if the train speed is too high or if the train is certain Diagnostic functions must not be suspended (for example, diagnostics loading profile for High-speed passenger trains).

box 224 ("DAQ timing set") is from box 223 Entering and corresponding to the setting of certain parameters that regulate the times of data transfer, such as the start time and the frequency of data transfer in the case of a number of measurement channels.

box 225 ("start SMI dta acquis.") is issued by 224 This is followed by a wait cycle (until the appropriate time has elapsed), followed by the start of the data transfer process for the sensors and instruments installed at the SMI. The process of box 225 is not necessarily performed by software, but only because it may correspond to the setting by software of a time or count in a "start register" of a corresponding hardware component (eg, a multi-channel signal generating device) comprising a number of clock and clocks Clock signals generated for data transfer processes.

box 226 ("SMI data acquisition") is through the process of box 225 It starts and corresponds to the acquisition of the measurement data of sensors and instruments installed on the SMI, which lasts until a specific message or signal is received, eg from a box 218 associated process. The topic of data transfer is used for the various instrument types that are involved in the SMI process can be installed in different sections of this document. The topic timing and synchronization of measurements is discussed in a special section (5.18) later in the text. The Applicant points out, as stated elsewhere in the document, that the method and system usually do not require a data transmission rate, since all or part of the measurement channels interfere with the vehicle displacement along the rails or with the vehicle speed (such as in Patent document [026]) and, on the other hand, that the relative timing accuracy of the various instruments is a major factor in ensuring adequate performance of the method and system presented herein. Data transfer rates for the sensors and instruments installed on the SMI are uniformly governed generally by performance-related principles, and they do not necessarily track the possible changes in train speed during vehicle sensing. On the other hand, a principle to be applied is to avoid unnecessarily fast data transfer, since this would lead to a waste of memory space and a slowing of the data processing, which would not be a significant improvement of the diagnostic services. More information on adjusting the data transfer rate to vehicle speeds can be found in various sections of this document, including the box below, in this section 218 associated processes.

box 227 ("close SMi data acquis.") is by Box 226 When the SMI train scan completes, it will execute a series of software instructions, such as closing measurement data files, setting indicators or parameters, and other possible actions such as starting data transfers between different data structures or starting a recording of Measurement data on stable data media.

box 228 ("set meas sleep mode") follows the exit from box 227 and corresponds to the start of a series of actions to establish the "stand-by status" of the system instruments, ie the box 220 Undo actions that have been executed.

box 219 follows the exit from box 228 , This is equivalent to completing a circle through the boxes 219 to 228 that correspond to the actions that are performed to take the appropriate measurement data for a move. The data processing tasks for the train, which after a short moment after entering the process of the box 226 usually started even after leaving the box 228 for the necessary time, which in principle depends on the features of the train, on the details of the methods of data processing used in the procedure and on the performance of the relevant data-processing equipment.

box 218 ("TDA &RTDP") corresponds to the box 232 from 3 and is with no other box of 4 connected because their processes in real time "in the background" of the cycle of Box 219 to 228 expire. In the stand-by status of Box 219 , corresponds to box 218 monitoring at least two wheel sensors at each of the two positions 206 and 207 from 2 (only one position, if the train transit on the corresponding track takes place in one direction only). A first wheel detection triggers the processes that exit the box 219 consequences. The processing of wheel release times for at least two wheel sensors is then performed in a more or less complex manner to estimate the value of train speed to the processes of boxes 223 and 224 and at an appropriate time to start the data transfer to the SMI through the function of the box 225 set. A very simple and often used way of achieving these goals is to calculate the train speed by dividing the distance between the wheel sensors by the time difference between the wheel identifications of the same wheel. The precision of the speed measurement usually depends on the characteristics of the wheel sensor, the data transfer characteristics, the distance between the sensors and the speed itself. One can obtain the average of more than one basic speed measurement for a pair of detection times to improve the measurement precision. A detailed description of the calculation of the vehicle displacement over time using the passing times of the wheels is given in section 0 below, but the corresponding process of Box 218 does not require the precision required for the "LDF function", as defined below: The possible grossest error in the estimation of train speed, the train position and negative accelerations, and an interruption in the speed estimate can be corrected by incrementing the last velocity estimate by a value from that time, from the most recently estimated velocity value, from the approximate time that the train takes to The grossest possible error in predicting the arrival time of the train front at the entrance to the area SMI can be estimated on the basis of the same for the speed estimated data. In practice, this error can be taken into account by starting the SMI data transfer at a sufficiently early time to prevent possible loss of the sample data. However, this aspect is not very critical, since the only significant consequence of starting the acquisition of SMI data too soon is a waste of disk space, which can be remedied later by deleting useless data. The problem of the worst possible error in predicting the arrival time of the train at the SMI and assuming the corresponding speed can only be significant if the distances 210 and 212 are much larger than the length of the shortest train to be tested, as a result of a relatively long time for the execution of the process by Box 220 started actions. A simple solution in this case is the installation of a single wheel sensor (one for each approach direction to the SMI) along the direction of travel to the location areas 206 and 207 that the start signal to leave the box 219 uses. The with box 218 Associated processes may also incorporate the data acquisition of wheel sensors installed at the SMI and perform real-time transit speed calculations to improve the data transfer rate while performing measurements. They may also generate alarm indicators and / or signals to indicate that the train is traveling slower or faster and below or above certain speed limits, making certain data channels unusable and rendering certain data processing tasks impractical. Another function can be with box 218 the sending of messages to indicate that the train end has left the SMI area, thus stopping the data transfer, ie the transition from Box 226 to box 227 is done (based on the wheel sensor signals and the speed estimate).

Professionals of the respective techniques like alternative schemes for achieving the goals of the above in the description of 4 to suggest addressed processes, such as the use of velocity prediction to program the timing of data acquisition, and the so-called "pre-trigger" data transfer technique to prevent wasting of measurement data, or starting the data transfer based on a signal for the data Arrival of a train, initially using a very high data transfer rate, which then decreases depending on the measurements made at the SMI.

In front the transition to the next Text section would like declare the applicant for the sake of simplicity, the majority of the explanation below assumes that the rails at the rail section at SMI straight are and indeed for a length on both sides, thanks to which one can regard the railway line as straight. The possibility an installation on a rail bend and the related Effects are explained in special section (0) below.

5.2 wheelset distances and the longitudinal position of vehicles over time

To Today, different types of wheel sensors (also wheel detectors or wheel trips) within different railway-related electronic counting systems of pulling axles, to detect the presence of a train, to measure vehicle speed, to associate [011, 017, 023, 024] wheels with vehicles distinguishing [025] from axles belonging to non-freight vehicles and for generating [021, 026] pattern time signals at a vehicle speed proportional frequency used.

The Time points to which vehicle wheels be recognized at certain positions along the railway track, can as explained below, used in the context of the procedure distances between wheelsets a vehicle and the longitudinal position to calculate a vehicle along the tracks over time. The values of the wheelset distances (here "WSD") are the basis for the Application of the explained below Vehicle identification method, and its accuracy is a major factor for the Efficiency of this method. The function for the representation of the longitudinal displacement of a vehicle (here "LDF") is a scalar function for time, the one fundamental importance for the functionality of the entire system, as it has within the procedures for allocation different types of measurement for a particular vehicle element has, as explained below is.

Two wheel sensors may be sufficient to measure the speed and direction of rail vehicles, and to easily calculate the distances between any two consecutive wheelsets if the train speed is constant or if high precision is not required for these measurements. A more complex set of wheel detectors and a robust and efficient computational procedure, on the other hand, are required in this case to ensure the required high precision in WSD determination and LDF, as the train speed when passing the train varies by the measuring interval SMI [017].

5.2.1 Relationships of WSD and LDS on the signals and positions of the wheel sensors

The composite graph of 5 shows how the Raderkennungszeiten relate to the vehicle movement, the distances between wheelsets and the positions of the wheel sensors along the rails. A typical freight car 275 with four wheelsets 280 - 283 belonging to two bogies is in the example of 5 used where a set of three wheel sensors 277 - 279 along the rails 284 is installed. The left diagram includes the wheel sensor signals 290 - 292 , each to the three wheel detectors 277 - 279 graphically with any arbitrary "offset" in any displacement units of the axis 290 over time in seconds from axis 285 are shown in the same way for the three graphs in this figure. The vehicle moves in the direction indicated by the arrow 276 with an initial speed of 30 km / h (at time 0 from axis 285 ) and with a strong negative acceleration (close to the maximum possible values in case of emergency braking). This low speed and the strong negative acceleration were chosen by the applicant to assess the effects of speed changes in the charts of 5 make visually clear. Ideally, the wheel detector signals correspond 290 - 292 from 5 a type of wheel detector with a two-state output that indicates a transition of the output when the wheel center is at a certain longitudinal distance from the center of the detector. The horizontal lines passing through the square pulses represent the times at which the positions of the center of the wheel coincide with the center of the detector. The increasing width of the wheel sensor pulses over time is the obvious consequence of the vehicle's negative acceleration. The axis 294 of the middle graph corresponds to the longitudinal distance in meters of the three wheel sensors from any position along the track rail. The direction of the axis 294 in this example is the same as the direction of the pulling movement. The symbols shown in the middle diagram correspond to the times at which a sensor detects a wheel center and the symbol shapes correspond to a specific wheel of the vehicle (eg the + symbol corresponds to the wheel 282 ). The first tip 291 and the graph point 292 correspond to the detection of a center of the front wheel 280 through the first sensor 277 It has happened while the following wheel recognition is on wheel 281 through sensor 277 relates and the point 293 in the middle graph. The values of the axis 289 in the right diagram correspond to the distance in meters, which has covered a car from time 0 on. Consequently, the axis shows 289 the longitudinal position of a particular vehicle point along the rail with the same direction as the axis 294 , The displacement of the axis 289 is the consequence of choosing the time 0 for the data 5 since this time the position overlap of the buffer front 295 and the wheel sensor 277 equivalent. The right graph is a graph of the LDF function indicating the LDF values read at the times when the vehicle is detected by any of the wheel sensors. The X-axis of the data shown in the right diagram of 5 is directly related to the distances 286 - 288 between the wheelsets of the vehicles, as shown in the three figures of the rail vehicle 275 is represented for the wheel center, which coincide with the LDF data. Each data series relative to a particular wheel sensor corresponds to the pattern of wheelset distances as the X-axis in the LDF diagram; eg the LDF data correspond 296 . 297 . 299 and 300 the wheel sensor 277 and the wheelset distances 286 - 288 , Conversely, the X-axis difference of the LDF graph for the data of a particular wheel corresponds to the distances of the corresponding wheel sensors; ie the difference on the X-axis of the data points 298 and 296 is equal to the distance between the wheel sensors 279 and 277 , The X-axis difference of any two points in the LDF graph is an algebraic combination of wheelset distances and distances between wheel sensors; ie the X-axis difference of value 301 and 298 corresponds to the distance 286 between the wheels 283 and 280 minus the distance between 279 and 278 , You can also see that the (right) LDF graph in 5 can be obtained from the left-hand center diagram by "shifting" the data triplets corresponding to a particular wheel, keeping the X-axis difference of the members of the triplets and the amount of displacement corresponding to a particular wheelset spacing. The methods for determining the wheelset distances and the LDF of wheel sensors and the wheel sensor distance data will be explained in more detail in this text. First, a brief discussion of wheel sensors and measurement errors in their use follows for the purposes of this invention.

5.2.2 Wheel sensors

The wheel sensors in the most commonly used class are based on electromagnetic sensing techniques. Some of the first devices of this class were based on a U-shaped magnet installed near the inner side of a rail and having at one of its poles a detection coil for detecting the signals; the wheel rim became over when passing the magnetic field recognized the sensor device. Later, excellent results were obtained with another type of electromagnetic sensor [022], which involves high recognition accuracy and fast response times by detecting the phase of radio frequency (RF) radiation radiated toward the flange and reflected by its metal surface. Honeywell's RDS80001 and RDS80002 track wheel sensors [950], as well as previous models such as the 926FS30-060-Z911-H, are widely used inductive-type high-speed proximity sensors with two-state dual-line current output

she have one High frequency generator (about 230 kHz) with an open magnetic circuit; the presence of the lane wreath in the sample area influences the respective magnetic Field and the resulting attenuation of the cycle fluctuation is registered by the sensor electronics. Another in the trade available Family of powerful high-speed sensors for railway wheel detection is the VRS series [951], developed by Invensys Sensor Systems / Electro Corporation (now owned by Honeywell). Recognize the VRS sensations the appearance and disappearance of an iron alloy box in the registration area by a permanent magnet and a sensing coil, because the resistance variation in a positive or negative peak the generated signal results. The VRS sensors are in many made of different versions and are particularly interesting for the process because of its high measurement bandwidth, which is usually over 15 kHz. Also different Models of Hall-effect sensors can be used by Honeywell as one Alternative with a high bandwidth to the o.g. VRS sensors to be delivered. A special kind of electromagnetic sensor is the "DRT Electronic Pedal "by General Electric Transportation Systems [963]. This device mounted two units on both sides of the rail and acts as a magnetic transmission and receiving device. Other excellent electromagnetic sensors used for assembly are suitable and approved on rails, and the wheel recognition of Vehicles are available from other manufacturers.

Other Types of sensors, and in particular certain fast-response optical sensors, also come for the implementation Contactless fast speed detection of wheel flange rims in question. Although optical sensors because of their assets, very precise Values for the wheel transit times are attractive, are electromagnetic Sensors for This application is especially recommended since it has been proven that with all the requirements given by the rail environment in this application are compatible and in particular with each kind of weather condition very reliable are inclusive in snow and ice without the need for special protection and heating systems. Practical mounting accessories are for everyone Commercially available electromagnetic track wheel sensors readily available, such as for example Honeywell RDS-CL-01 "Rail Mount" No rail drilling required and easy along the rails can be positioned.

5.2.3 Measurement uncertainties when using wheel sensors

The descriptions in this document make it clear that the implementation of the method and system does not impose strict minimum requirements for the accuracy, resolution, response time, or other performance of wheel sensors. However, this document does explain how uncertainty about timing and position for wheel recognition adversely affect some of the system's performance. It is therefore worthwhile to investigate the causes of such measurement uncertainties or errors, at least for the electromagnetic wheel sensors, so that the choice of the wheel sensors in the implementation of the system is easier. Usually, electromagnetic wheel sensors (except for the above-mentioned GET DRT Electronic Pedal) are mounted on the inner rail side just below the rail head to detect the wheel rim when interfering with the electromagnetic sensor field. Non-cylindrical symmetrical wheel wear may lead to a delay or delay in the time at which the sensor indicates the arrival and departure of the wheel center. Such a kind of error is more or less critical depending on the type and model of the wheel sensors. A major factor in determining uncertainty of measurement, especially for large diameter and relatively high speed wheels, is sensor noise, while a sensitivity shift over time or for other reasons such as temperature is not important if the sensor signal is as described below specified instructions. The machining method given below also compensates for the effect of a variable lateral position of an axis perpendicular to the rails, which can symmetrically change the times (arrival and onward travel of the wheel) at which the sensor registers a certain level of disturbance in its electromagnetic field. The yaw of wheelsets, which leads to a Radschlingern (see wheel position in 6 ), however, may cause an error due to the change in the sensor response at various distances from the rail edge of the leading and trailing wheel halves of the wheel. Another source of uncertainty, which may be more or less important, depending on the type of wheel sensor, and which usually involves the wheel detectors with a two-state output is the hysteresis that can cause various changes in the arrival and departure time of the wheel. The extent to which the hysteresis affects the determination of the time at which the wheel center passes depends strongly on the wheel diameter and may also depend on the lateral displacement of the wheel center from the rail axis. The hysteresis effect can be partially corrected by calibration and by taking into account the wheel diameter. Finally, the detectors for which the output is generated by a digital system such as a microprocessor or a DSP can produce a significant random error associated with signal probing during the separate time intervals between sensor output updates. Of course, the time uncertainty also depends on the data transfer system, which is discussed in more detail in this document. A quantitative estimate of wheel transit transit times, the uncertainty associated with the aspects described above, and other causes of failure are discussed later in the document.

5.2.4 Measurement uncertainties in the context of the yaw fluctuation of wheelsets

6 refers to a known feature of wheel set kinematics that must be considered in the design of the system and in the processing of wheel sensor data when very accurate measurements of wheelset distances and vehicle positions are desired. In general, wheelsets of rolling rail vehicles do not move over the railroad tracks simply by running their roll axle vertically over the rails centered over the rail axis 322 stay. Especially mounted on the bogie wheelsets are often a heavy movement about the vertical axis 321 exposed, which in some cases leads to a lurching of the wheel. 6 shows a bogie with two axes close to the limit of its heavy vibration with the extent of the pendulum motion 320 close to the maximum value and a wheel 316 that the rails 311 attacks. The wheel 315 on the left rail 310 lies in the transit direction 317 around the length 318 further ahead as a wheel 316 on the right rail 311 , This means the wheels 315 and 316 each in the longitudinal direction positive or negative 319 are shifted in relation to their common axis. The range of yaw variation of bogies varies widely and depends in a complex manner on a variety of factors, but the approximate maximum value of the yaw or pendulum angle 320 , and thus the shift differences in the longitudinal direction 319 and 318 , can be easily calculated based on the gauge line profile 325 (whose actual maximum value takes into account deformations and rail wear), the distance between the inner side of the flange, the minimum flange thickness (taking into account the maximum flange wear) and the separation distance between the two axles. Bogies with two axles with smaller wheels and smaller interaxle distances have larger displacement differences 319 and 318 on. In practice, a shift difference 319 15 mm larger and results in a 30 mm greater change in the immediate distance between the wheels belonging to different bogies.

However, the influence of the Yaw angle on the longitudinal position of the wheel center can be neglected. Therefore, the wheelset centers are more suitable for accurate estimation of vehicle displacement and static or average distances between wheelsets than individual wheel centers on a rail. The use of wheel detector pairs like 312 and 313 in 6 allows the calculation of the time at which the wheel sets pass a certain position from the transit times of the two corresponding wheel centers. The two quota lines 323 and 324 refer to the "wheel detector centers", ie the longitudinal position along the rail at which registration of the wheel takes place. The difference 314 between such quotas 323 and 324 as a result of the actual installation of the sensors may be large enough to be used in the calculation of the transit times of the wheelset centers, as explained below.

5.2.5 WSD and LDF off the transit times of wheelsets to calculate

With reference to the above comments too 5 The following is the problem of calculating the distances between a set of J wheelsets from the transit time values generated by the wheel sensors or wheel sensor pairs K positioned at known points along the rail, wherein possible speed changes of the rail vehicles during the measurements the system will be considered.

The values K · J, time values t indicated by t _{j, k} , are defined as the times at which the wheelset center j passes the height of the registration center k along the rail. The L _{k of} each value K is the distance from the registration center k from a reference position along the rail, where L is a longitudinal axis parallel to the rails. With λ _j , the coordinate of Radsatzmitte j is indicated on the longitudinal axis Λ, which is integral with is the center of a certain wheel center j0 and whose direction is by definition opposite to the train transit direction and to which the value s is associated, where s is 1 or -1, depending on whether this direction is equal to or opposite to the direction of the axis L. The wheelset center j0 was arbitrarily set as zero of the axis Λ. The sought LDF function of time is given by L (T) and is defined as the coordinate of the starting point of the axis Λ, ie the wheelset center j0, on the axis L at time t.

The set of K · J equations L (t j, k ) = L k + s · λ j , with j = 1 to J and k = 1 to K (100) would apply to a series of J wheelsets mounted on a completely rigid (no relative play of wheelset centers) vehicle or a rigid set of vehicles, and if no uncertainty had to be taken into account for the values L _k and to t _{j, k} . Even in such an ideal case, the only recognizable vehicle positions according to the set of equations 100 would be those which simply correspond to the values of the K variables L (t _{j0, k} ) which are equal to the corresponding value L _k since λ _{j0 has been chosen} equal to zero , The other J-1 values of λ _j and K (J-1) values of L (t _{j, k} ) remain indeterminate since the input data for the problem does not provide information about the speed of the wheelsets between two successive transit time measurements a wheel group included. The calculation problem is thus defined as the search for a mathematical LDR function L (t) approaching the true value LDF, taking into account the limits of the vehicle kinematics, which are the maximum values of the acceleration and the negative acceleration and of the maximum rates and time intervals on the basis of which the negative acceleration and acceleration of actual rail vehicles can vary over time in practice. In addition, the uncertainties of the input data and the presence of mechanical play should also be considered in the search for a suitable LDF function.

durch Ändern der Parameter der Annäherungsfunktion L(t) und der Werte J – 1 von λ_j für j ≠ j0.By neglecting the uncertainty of different input data, the LDR problem can be reduced to a function approximation to the minimum squares based on minimizing the amount

by changing the parameters of the approximation function L (t) and the values J-1 of λ _j for j ≠ j0.

die eine Chi-Quadrat-Größe der Richtigkeit der Näherung darstellt. Die Gleichung 102 ist geeignet, wenn die durch die Änderungswerte σ 2 / j,k ausgedrückten Ungewissheiten einer Fehlerverteilungskurve folgen. Andere Gleichungen könnten verwendet werden, wenn zum Beispiel eine wichtige Ungewissheitskomponente einer Quadratverteilung folgt, wie im Falle von Zeitfehlern im Zusammenhang mit getrennten Abtastungen.In the case of the LDF problem, various uncertain values can generally occur on different input data, for reasons explained later in this document, and so it is better to minimize another amount for the LDF approximation function

which represents a chi-squared size of the correctness of the approximation. The equation 102 is suitable when the uncertainties expressed by the change values σ 2 / j, k follow an error distribution curve. Other equations could be used if, for example, an important uncertainty component follows a square distribution, as in the case of timing errors associated with separate samples.

The choice of mathematical expression for L (t) is of course critical because the chosen function must be flexible enough to accommodate the input data without having unrealistic behavior in the distances between the values t _{j, k of} the independent variable t. In this case, the above-mentioned physical constraints on vehicle dynamics may be used to limit the least squares minimization, such as minimum and maximum values for the second and third derivatives of L (t), or addition to the correct equalizer 101 or 102 of FIG one or more members whose minimization corrects the approximation curve.

One particular type of curve that Applicants consider appropriate for the LDF approximation is a spline function S (t), ie a piecewise function of cubic polynomials S p (t) = a p (t - t p ) 3 + b p (t - t p ) 2 + c p (t - t p ) + d p (103) in which p = 1 for P and t _p indicates the lower limit of the time domain of the polynomial p, while the upper limit is equal to t _{p + 1} for p <P.

According to the standard spline definition, the P-1 functional continuity conditions become S p (t p + 1 ) = S p + 1 (t p + 1 ) (104) set along with the P-1 continuity conditions for the first derivative over time F p (t p + 1 ) = Ṡ p + 1 (t p + 1 ) (105) and P-1 continuity conditions for the second derivative over time S .. p (t p + 1 ) = S .. p + 1 (t p + 1 ) (106) where p = 1 for P-1 for each of these sets of equations. Equations 104, 105 and 106 are each written in three sets of equations a p (t p + 1 - t p ) 3 + b p (t p + 1 - t p ) 2 + c p (t p + 1 - t p ) + d p = d p + 1 , (107) 3a p (t p + 1 - t p ) 2 + 2b p (t p + 1 - t p ) + c p = c p + 1 (108) and 3a p (t p + 1 - t p ) + b p = b p + 1 , (109) where p = 1 for P-1.

The spline S (t) is defined so far by the 4P parameters a _p , b _p , c _p and d _p and by a set of 3P-3 conditions. The missing conditions can be covered by constraints if the input data is to be interpolated, and also by constraints on the derivatives of the outer values of S (t) and / or other conditions to minimize the variations of the splines between the data points.

A common problem with spline functions is the fluctuation of the curve in the area of the fitted points, especially when P is close to the number of value that needs to be adjusted. One possible control of this difficulty is the minimization of W, defined by the equation W = ρ · x 2 + ∫ | S .. (t) | 2 dt, (110) wherein the scalar ρ determines the importance of the x ² minimization defined by equation 102 over the minimization of the integral term that acts as a weakening element of the curve.

In the case of the problem of the LDF approximation, the flexibility required for the approach curve depends on the train speed, since the time passing between the lowest and highest values t _{j, k} possible changes in acceleration and negative acceleration limits. So, if a spline function is used, P should be reduced if the average train speed is increased during the sensor measurements. The number of applied wheel sensors or wheel sensor pairs and the corresponding uncertainties in the measurements of the wheel transit times as well as their position on the rails can also be taken into account in the choice of P.

The time intervals between two consecutive values of t _{j, k} are not constant, since they depend on the values J-1 of λ _j , while the probability of certain acceleration changes depends exclusively on the duration of the time interval considered here. Therefore, Applicant proposes that the time domains of the P polynomials have the same duration, and the P-1 values of t _p for p range from 2 to P and are equidistant from the minimum and maximum values of t _{j, k} , while t ₁ is t _1,1 and t _{JK is} equal to the maximum time value from the range S _p (t). Another possibility is to use adaptive values for t _p for p that range from 2 to P and are subjected to the minimization process, and this choice allows the use of a relatively small value P. The actual fit, and especially that defined above Spline function with respect to the corresponding condition sets 107, 108, and 109, by minimizing W, as defined in Equation 110, and by using additional constraints on derivative values, is not discussed here, as it is done using standard techniques and algorithms can be performed, which are described in existing literary publications. Although the choice of P and ρ is based on theoretical considerations, one should optimize two such numbers as a function of vehicle speed as the basis of the experiment.

5.2.6 Calculation of the values t _{j, k} and σ _{j, k}

When an ideal wheel rolls at a constant speed and its transit time is measured by an ideal sensor which transmits its trip signals symmetrically with respect to the distance of the wheel center from the sensor center, then the wheel center t _C release time is just the average between the first and last Tripping signals, namely t _L and t _T.

For acceleration or negative acceleration, the use of equation 111 introduces either a negative or a positive error in the estimate of t _C , t _L and t _T. This error is greater with absolute values of acceleration and negative acceleration, with low speed values and with higher distance values L _L - L _T between the longitudinal positions of the points where the wheel sensor generates the first and last triggering signals. The error can be easily corrected by first calculating the value t _C using equation 111, performing the LDF calculation, and then using the acceleration and velocity values from LDR for a recalculation of t _C , which is finally used for a more accurate calculation of LDF and WSD is used. Simple kinematic calculations show that very strong braking (the worst case absolute value of negative or negative acceleration) with a value L _L - L _T of 300 mm results in an error in the estimate of t _C , expressed as a corresponding error the position of the wheel center is less than 1 mm at a speed of 15 km / h, and drops to 5 mm in the case of slightly less than 6 km / h (the dependence of this error on the speed is not linear). From this one can conclude that the above correction is only worthwhile if very precise wheel sensors are used and if the measurements are made at very low pulling speed. If the wheel sensors selected for the system have an analogue output, the tripping times t _L and t _{T must} first be calculated, using a numerical method that matches the characteristics of the sensor and takes into account linearity and noise.

in der t_LWL und t_LWT für die ersten und letzten Auslösezeiten des Sensors des linken Rads stehen, und t_RWL und t_RWT für die des rechten. Die Verwendung der Gleichung 112 impliziert offensichtlich, dass der entsprechende Wert von L_k sich auf die durchschnittliche Längsposition entlang der Schiene des Sensors für das linke Rad bzw. für das rechte Rad bezieht. Auch in diesem Fall kann gezeigt werden, dass der Fehler im Zusammenhang mit der Vernachlässigung der negativen oder positiven Beschleunigung bei größeren Geschwindigkeitswerten vernachlässigt werden kann. Eine Korrektur dieses Fehlers kann auf die oben für einzelne Radsensoren beschriebene Art und Weite korrigiert werden, wobei man in diesem Fall die Positionen der beiden Radsensormitten als Paar betrachten muss, was erheblich anders sein kann, wie oben im Kommentar zu 6 bemerkt wurde.If pairs of wheel sensors are used as described above to compensate for the yaw of the bogie, and if the acceleration is neglected, then the transit time t _{WSC of} a wheel center can be calculated by the following equation

where t is _LWL and t is _LWT for the first and last tripping times of the left wheel sensor, and t _RWL and t _{RWT are} for the right wheel. The use of Equation 112 obviously implies that the corresponding value of L _k refers to the average longitudinal position along the rail of the left wheel sensor and the right wheel sensor, respectively. Also in this case it can be shown that the error associated with neglecting the negative or positive acceleration can be neglected at higher speed values. Correction of this error can be corrected to the type and width described above for individual wheel sensors, in which case the positions of the two wheel sensor elements must be considered as a pair, which can be significantly different, as discussed in the comments above 6 was noticed.

The values of L _k should be precise and consistent with the measurement errors of the wheel sensors. The L _k values refer to the "detection centers" of the sensors, which do not necessarily coincide with the apparent center of the housing. Depending on the type of wheel sensor, it may be useful to measure the L _k values by observing a controlled movement of an iron box compared to the position along the rails. However, the measurement of L _k values should be part of the general calibration method of the system, which depends on various technical options for sensors, mechanical support structures, data acquisition system and other design elements of the actual system to be used.

The uncertainties of position and timing to which the wheelset transit relates play a role in the process, as has already become clear in the above discussion of Equation 102, and as explained in other parts of this document. The importance of the individual sources of error in the implementation depends on the decisions on the type and model of the wheel sensors, and must therefore be analyzed and taken into account in the design and development of the applicable system version. In any case, it is worth noting that some of the relevant measurement or calculation errors are more correctly defined as position errors, and others than time errors. The values of σ _{j, k} in Equation 102 relate only to L (t) and not directly to t. Although the LDF approximation curve can be performed by explicitly using separate estimates of space and time errors in the corresponding equations, it might be convenient to use, for example, in Equation 102, σ _{j, k} values that are the appropriate combination all relevant time and length errors are after converting the time errors to length errors by multiplying them by the speed value. In this case, the speed value to be used may be an approximate value calculated by finite differences based on the values t _jk and L _k and together with the equation 100. Alternatively, the LDF calculation may be repeated after calculating σ _{j, k} using the speed values of the first LDF calculation. Another possibility is the tabulation of σ _{j, k} as a velocity function.

Applicant emphasizes that in general many variations are possible for the LDF calculations, and most importantly, that the above-proposed approach of the LDF approximation for specifying the input values of t _{j, k} and σ _{j, k can} be avoided if one put the appropriate calculation expressions into the formula for the value that needs to be minimized for the LDF approximation. The demonstration of the best performance of the best wheel sensors in terms of implementing the system, proper design of the data collection system and the correct installation of all the important components of the system make missing wheel identifications or incorrect wheel recognition very rare. However, it is possible to implement a diagnostic algorithm to identify missing or incorrect wheel sensor signals. False signals would be eliminated. For missing radar recognition times one could delete the corresponding time from the expression of the amount to be minimized for the determination of LDF.

5.2.7 Interpolation and extrapolation errors in the positioning of wheel sensors

The number and locations of the wheel sensors mounted in the system's range are not strictly defined in this document and may vary within certain limits, depending on the performance of the system implementation and the choice of particular system components, including the wheel sensors themselves. The drawings off 7a to 7h This document is used below to discuss the positioning of the wheel sensors in conjunction with the consequent effects on the calculation of WSD and LDF as well as the precision in associating different measurements with vehicle parts, as further explained in the document below.

A distinction between interpolated and extrapolated LDF values helps to analyze the question of sensor positioning. The above-mentioned LDF calculation provides a function L (t) approximately indicative of the time value and the value of the distances of the longitudinal displacements between measured points (or between the times of wheel detection). The LDF estimate within these intervals is interpolation, and the precision of a particular LDF value in these intervals depends on the precision of the measurements, the values for that particular average speed interval, the specific interval width, and the distance of a particular interval define such point from the interval limits. Also, the values and uncertainties of other data points near the extreme ends of the interval may affect the accuracy of the LDF estimate within the interval of interest. Finally, the details of the ZDF calculations (eg, the choice of interpolation function, the convergence criteria, etc.) have an obvious impact on the interpolated LDF data. The estimate of the LDF values for a point in a range t or range L that is outside the interval of the value used for the LDF calculation is an extrapolation of the L (t) curve and is subject to increasing uncertainty because of the greater separation of time and space than in the last time and position measurement of the wheel. The extrapolation would normally be performed by extending the interpolated LDF with a smooth curve and imposing continuity at the intersection point with the interpolation interval on the curve and its first or first and second derivatives. However, it is clear that the extrapolated estimate and the actual LDF may differ significantly as the acceleration changes outside the LDF definition interval.

The dashed area 403 shows from 7a to 7h the part of the measuring interval "XSMI" of the system of length D _XSMI , which is defined in the same way as SMI, but _excluding the wheel sensors. The same is true in all these illustrations 405 for a generic rail vehicle with two articulated bogies while 406 . 424 and 423 stand for a rail vehicle with a joint bogie. The direction of the rail vehicle is indicated by a thick arrow above the car body. For the sake of simplicity, a constant distance D _WS between the wheelset sensors or between sensor pairs is assumed. The interval along the rails between the first and last wheel sensors is called "WSI" (for Wheel Sensors Interval). The length of WSI is D _WSI = L _K - L ₁ . The union of XSMI and WSI obviously corresponds to SMI. Cases are considered (eg in 7a ) in which XSMI exceeds the WSI, and others in which the ratio is reversed.

7a and 7b refer to the estimation of the maximum width D _{MII of} the LDF interpolation intervals in the case where the distance D _IUB passing through 401 between the closest wheelsets 427 and 428 is displayed on each of the two articulated bogies of the rail vehicle 405 greater than D _WSI + D _WS . The largest interpolation interval length D _MII corresponds 400 , This largest interpolation interval commences in accordance with the time (on which 7a and 7b refer), on which the wheel group 427 the sensor 404 leaves and ends when the wheel 428 the sensor 429 reached. The only practical difference between 7a and 7b That is, the LDF is used at the largest interpolation interval to make measurements to a greater or lesser extent than the carbody.

7c refers to a case where a vehicle 405 achieved with two articulated bogies XSMI. At the time according to 7c the first vehicle is detected and thus the XSMI measurements for the length 408 the vehicle body of the LDF extrapolation, if only the wheels of this vehicle are used for the approximation of L (t). The length 408 is clearly the sum of the distance 410 between the limits of XSMI and WSI plus the distance 409 between the first wheel group of the vehicle and the front edge of the vehicle itself. 7d similar 7c but in this case the vehicle leaves the measurement interval XSMI. XSMI measurements of the vehicle body refer to the extrapolated LDF values for the length 411 of the vehicle and this length is the sum of the distance 413 between the limits of XSMI and WSI plus the distance 412 between the last wheel group of the vehicle and the rear edge of the vehicle itself. 7e and 7f refer to the conditions where the interpolation interval can be set equal to zero, if necessary, in the cases where the vehicle is scanned with two articulated bogies and only the wheel sensor measurements for the own wheelsets are used for the LDF calculation. 7e refers to a vehicle approaching the XSMI, showing that the unused interpolation length 415 is equal to the difference between the distance between the XSMI and WSI boundaries and the distance between the front wheel group and the front vehicle edge. In a similar way shows 7f in that for a departing vehicle whose scan is completed, the unused interpolation length 418 equal to the difference between the lengths 417 and 419 is the 416 and 414 in 7e correspond.

Of the first reason why the above cases are being considered were in which only the wheelsets of the Vehicle itself for the corresponding LDF approximation used is that a single vehicle, like one Locomotive without attached Cart through which the system can be scanned. A second reason is that all trains have a first and a last car, for which the above observations be valid. A third reason, the LDF calculation only on the wheelsets of a Vehicle itself could be based the game in longitudinal distance two adjacent vehicles, also called buffering. The buffering can be much larger than the long game between any two sets of wheels that are different Bogies belonging to the same vehicle and so supplies the use of wheel sensor data only the on-board axles usually one more accurate LDF estimate as the use of own axles together with axles of neighboring vehicles.

There are cases, like the ones on which 7g in which the avoidance of a strong Redu Extrapolation when using only separate wheelsets for the ZDF calculation of a vehicle to a very large extent 422 from the WSI. In the case of a rail vehicle 406 equipped with an articulated bogie 430 , would the zeroing of the distance 420 to avoid extrapolation at a distance 422 from the end of the XSMI to the last wheel sensor, equal to the distance from the rear edge of the vehicle body 406 to the second wheel of his articulated vehicle frame 430 is. In practice, the situation is of 7g in principle applicable to semitrailers on bogies, as these represent the main class of articulated rail vehicles, since it can be assumed that for them the monitoring of the boundary line profile and overheating must be relevant. The importance of buffering in the particular case of semitrailers on bogies is fortunately extremely low, thanks to the special design of couplings between vehicle bodies and their bogies. Therefore, the applicant proposes that the case of 7g to a case of interpolation as in 7a is reduced, and, with respect to 7h , the wheelsets of bogie 425 of the vehicle 423 as well as bogie 426 from vehicle 424 for the calculation of LDF for vehicle 423 to take into account.

in the Generally, the system is for the scanning of any kind of train moving in each direction, and therefore WSI and XSMI share their centers along the rail and thus offers a symmetrical situation, what the distance the WSI and XSMI edges at each input / output of SMI.

The actual D _XSMI results from the choice of sensors positioned on the SWI for vehicle _scanning , as explained later in the document. _{As discussed} above, a large value D _{XSMI may imply} a similar value D _WSI , but a very short _XSMI does not reduce the D _XSMI below a value that depends on the largest distance between two wheelsets used for the LDF calculation in Be considered.

Once the SMI is selected in the design of the system implementation, the values for D _XSMI and the number of wheel sensors or the value for D _{WS can be selected} depending on the desired extrapolation _length and the uncertainty of the transit times of wheelset centers. This choice can be made using kinematics based calculations, or it can result from a comprehensive numerical simulation that takes into account different train shapes and the most negative case values or statistics on the worst-case values of positive and negative acceleration over time.

5.3 Additional Measurements in vehicle identification

The Vehicle identification procedure, described below is used as a fundamental input the WSD records that can be generated together with the LDF calculation, but it can be the availability use certain other measurements and processing techniques, the quick and easy detection of other vehicle features enable and help with the choice of the "candidate model".

8a . 8b and 8c relate to the use of a non-contact optical sensor for measuring the approximate diameter of wheels, and to obtain information about the shape of a wheel body. Below, some additional applications of this measurement within the system are described. The optical distance measuring device 350 in 8a and 8b for creating wheel profiles has a laser beam 353 which hits the measurement target at a point from which a portion of the laser beam becomes the instrument 350 reflects, which measures the angle of impact.

Some of LMI Selcom's Optocator ^TM [952, 954] faster laser rangefinders are particularly suited to this application, thanks to measurement bandwidth, viewing distance, test beam diameter, noise level and precision, and have already been proposed for the measurement of wheel wear on rail vehicles used [027]. From this series of laser triangulation devices, the distance is measured by a phase-coherent demodulation of the detected laser beams. Fast feedback based on the intensity of the laser radiation back to the detector allows the sensor to operate with a wide range of retroreflective values for the measurement target. In particular, the model Optocator ^™ 2008-180 / 390-B (part # 813214), a laser distance sensor, has a measurement range of 180 mm, a standstill distance of 390 mm, a test rate of 62.5 kHz with a bandwidth of 20 kHz, 0.28 mm RMS noise, ± 0.2 mm precision, laser dot size 0.65 mm and IP65 packaging. Among the many alternative models, the Optocator ^™ 2207-200 / 325-K (Part # 809516) has a 200mm range, 325mm installation distance, a 32kHz test rate with a bandwidth of 10kHz, 0.3mm RMS noise , ± 0.4 mm precision and a laser spot size of about 3 mm. As a third example for a model with a larger installation distance, the model Optocator ^™ 2008-400 / 1178-B (part # 809451) is called; This laser distance meter has a measuring range of 400 mm, an installation distance of 1178 mm, a test rate of 62.5 kHz with a bandwidth of 2 kHz, 0.5 mm RMS noise. Several other potentially suitable Optocator ^™ models are available and the final choice of sensor should be made taking into account the installation geometry discussed below and the desired performance with respect to the resulting profile quality.

Applicant prefers to install the laser distance meter on a rigid mechanical support fixed to a very stable mechanical frame with a solid foundation, rather than on the sleepers 348 or the rail 347 to fix so as to avoid problems associated with intense vibration or shock.

The angle choice 355 between measuring beam and a plane 354 parallel to the rail head 347 due to limitations of the gauge line profile (for the lower vehicle area) affects the minimum usable standpoint distance and determines the inclination of the measuring area that the 358 defined over the wheel area. A special and attractive slope corresponds to an angle value of zero 355 , when using a larger installation distance than in 8a , It is also possible to use a sensor with a large installation distance while working with an inclination of the laser beam whose zero point of the beam is higher than the measuring point.

Another installation parameter that must be chosen is the approximate height above the rail head plane where the laser beam strikes the wheel, taking into account the measurement target. The choice of the sensor model and the angle 355 determine the possible sensor positioning, depending on sensor size or size of an additional housing, to protect the sensor from environmental influences and from damage caused by falling stones from the tensile load 349 might be necessary. A higher angle value 355 makes the protection of the sensor optics more difficult.

you can orient the sensor so that the laser beam is not vertical to the rail direction, but this would lead to the loss of the profile measurement symmetry to lead, and make data processing more complicated. Such a turn may be justified, however, if the measurements are used to evaluate the wear of the wheel profile, as in the case of [027].

A protective cover that opens automatically when the train arrives is recommended. Flushing the sensor or the protective housing with air may be useful to keep the optics clean and, if necessary, help extend the ambient temperature compatibility range. An active system that keeps the sensor temperature within a narrow range is not necessary for this application when using Optocator ^™ sensors, as these have a maximum temperature deviation of about 100 ppm of the measurement range per degree Celsius and, if desired, those below described deviation compensation can be applied. However, it is recommended to provide a heater for de-icing the sensor housing if the sensor is to be installed in a location where snow and ice may be present.

LMI Selcom Optocator ^TM devices use pulsed laser diodes with different NIR or visible wavelengths and their power is large enough to take laser safety precautions according to current regulations. Some of them have a maximum pulse power of 20 mW at 780 nm with a pulse duration of 32 μs and belong to laser safety class 3B, according to standard EN60825 (1991). Installation on positions proposed in this document for wheel profiling does not directly expose rail travelers and crew members of a train to the laser beam, but laser safety measures such as blocking the laser energy in the presence of a passing train and installing a radiation blocker in one suitable position depending on the angle 355 might be necessary.

The data is quickly output from the Optocator ^TM sensors in digital form with a proprietary data standard based on the same electrical characteristics as RS422. A special double-height VME board [953] is available from LMI Selcom to read the sensor measurements in real-time, allowing data transfer and measurement synchronization within a standard VME rate. Additional output signals that integrate with non-VME bus devices such as PC data cards or PLC units are also available.

The curves 362 and 363 represent two simulated wheel profiles in a graph in which 360 the measured distance and 361 represents the time. The steep rise 364 the wheel profile curve corresponds to the over from outside the registration area (or reading a target in the background) for the first reading of the lane rim 346 , The following curved part of the signal corresponds to the reading of the flange side and the wheel rolling surface 345 until the outer wheel surface is measured. The profile 363 corresponds to the example of a particular wheel with a completely flat wheel center, at least in the area that is close to the wheel tire. The rolling practically does not affect the wheel body profile, unless the wheel body is not symmetrical to the roll axis, as in the case of "corrugated" wheel bodies.

The first step in processing the profile data is to replace the time domain with the longitudinal displacement range by applying the appropriate LDF, which will be particularly accurate when the wheel profile sensor is mounted close to a wheel sensor. The approximate wheel diameter or radius can be determined by approximating the values of the wheel profile and making a simple geometric estimate assuming a certain approximate height of the railhead. The estimation of the wheel diameter is subject to various factors of uncertainty, such as the wear of the wheel profile, the rail head profile, the irregularities of the roundness of the wheel, eg due to flat spots, but these factors do not lead to an oversized diameter classification in the vehicle classification procedure discussed below Error. By contrast, a significant systematic error results from the lowering of the rail head due to wear, galling and ballast deformation. Therefore, it is necessary to update the value of the rail head height in one of the few ways n. The first option is to measure the rail head height with a laser distance meter installed on the SMI for the determination of the limit line data. A second option is to install all standards. A third possibility is to position a beveled target on the rail on the inner rail side and to perform a measurement of the background with the wheel profile sensor, provided it is at a compatible angle 355 Installed. Two or more of these techniques may also be used together to improve the precision and reliability of the wheel diameter measurement function.

One alternative way for the calculation of the wheel diameter without the impact of The time-reducing railhead is the installation of two fast laser distance sensors whose measurement levels are at the wheels at two different heights above the Cross rolling surface. In this way, an assumption regarding the height of the rolling surface is not necessary. Lowering this area can also result as a result of wheel diameter calculation.

The expected over time lowering of the rail head should also be taken into account when looking at the angle 355 and the height at which the profile sensor is installed. Thermal expansion of the corresponding structures can thus lead to a slow change in the position of the Radprofilsensors and above all to a tolerance shift. A fixed target beyond the rail, at a reasonable angle 355 , or a Abstandseichklappe in front of the rail can be used to compensate for the tolerance shift.

Using the fastest test rates and the highest bandwidth available in the Optocator ^™ range allows the identification of high quality profiles for the purpose of this application. Example: a 62.5 kHz test rate corresponds to a distance between the measurements of about 0.5 mm and a 20 kHz bandwidth means a reaction time of 16 μs, which also corresponds to about 0.5 mm with a 90% signal separation after an abrupt change in distance. In each type of these sensors, the peak-to-peak noise varies with the square root of the bandwidth; Applicant's simulations based on the manufacturer's Optocator ^™ specifications have shown that data approximation allows satisfactory profiles to be determined in terms of edge detection accuracy.

Although this document only discusses the fast LMI Selcom Optocator ^TM laser distance meters for wheel profile measurement, other types of profilometers can be used, such as the optical measurement systems described in [028] or a fast laser rangefinder such as a Zoller + device Frolich [961], which is discussed below in the context of measuring car body boundary line profiles. As stated in the text, it does not limit the use of wheel profile measurements to vehicle identification. In the context of the above wheel failure detection errors, the wheel diameter can also be used within the system for correction of wheel sensor data with respect to hysteresis. The measurement of the absolute lateral position of the wheels from the corresponding wheel profiles can be used to improve the estimation of abnormal thermal conditions in axle-related components described below in the document, as long as the relevant thermal radiation scanner is installed close enough to the wheel profile detector.

Installing a fast profile sensor like LMI Selcom Optocator ^TM in a similar way as in 8a and 8b at a height and at an angle suitable for making measurements at the average center height of the buffering of rail vehicles, may provide a signal that may serve to accurately determine the longitudinal positions of the buffering interface. Spacing profile laser equipment can also be installed at a different height to provide an accurate data set that can be used to detect bogies or other features of rail vehicles.

Wheel profile sensors such as LMI Selcom Optocator ^™ units could also be used as wheel sensors with excellent results, but applicant attaches great importance to restricting their installation to one or two aspects of wheel profiling and possibly the identification of other lower-vehicle profiles, above all because of the cost and because of certain installation aspects.

5.4 Primary vehicle identification procedure

9 Figure 12 shows a simplified flow chart of a software application that enables rapid and efficient identification of rail vehicles without requiring any sort of active or passive marking of any vehicle, and without information about train formation from other rail information systems. This application corresponds to the "primary vehicle identification" according to 234 in 3 ,

The primary vehicle identification (here "VI") can be found at the inbox 369 ("BEGIN") begin as soon as some data of the wheelset sensors, which correspond for example to about 20 m of the train, were adopted and transmitted in the form of a suitable data structure of the software of the VI procedure itself for read access. The data of the wheel sensors can be prepared in different ways and taken up by different techniques. In any case, they encode the wheel sensor data (here "WTD") in time format associated with a particular wheel sensor or pair of wheel sensors and a serial number corresponding to the wheelsets in the order in which they pass through the corresponding sensors or Sensor pairs were detected. In the discussion of 9 It has been assumed that the reading and management of system-acquired data is done by other processes that are addressed elsewhere in the document, and that other processes are used to generate flags indicating the end of the train scan, or exceptions, as the stopping of the train or the driving at a below the minimum limit driving speed by the SMI, are responsible. Canceling the procedure due to abnormal events or diagnostics alerts, as well as the actions to be taken on errors, are in the comments below 9 not addressed.

The most important VI application can generally be considered as a procedure that assigns progressively recognized wheelsets to vehicles that are defined and eventually classified as identified or unidentified. The LDF and WSD of such vehicles are calculated and used in the primary VI procedure and represent an output used by other applications in the system. The term "previous vehicle" (here "PV" for previous vehicle) refers to the vehicle to which the VI procedure last assigned a wheelset. box 371 ("is PV an IV?") refers to a branch depending on whether the PV is an identified or unidentified vehicle. In the special case where the VI procedure of Box 369 is started, there is no previous vehicle and the application continues to Box 373 ("get WTD"). On the other hand, if a previous vehicle exists and has been identified, the data is moved to the location of its buffers, if applicable, at step 372 ("get PV BID") queried from the vehicle database ("BID" stands for "Buffers Information Data"). Such BIDs preferably include at least the longitudinal coordinates based on the vehicle's buffer boundaries.

step 373 corresponds to reading the wheel transit time data for a series of wheelsets. A favorable choice is the WTD reading for two unassigned wheelsets and for the two previous assigned wheel sets, obviously excluding the first vehicle.

However, one can adapt the procedure of the first WTD read for a different number of wheelsets. Of course, the WTD reading depends on the availability of data from wheel sensors, and therefore, if necessary, a wait cycle is switched until the data is available or until an exception flag is issued. The applicant clarifies that the term "reading" for box 373 and others in the comment from 9 said boxing does not strictly express the transmission of data to the application, but, more generally, that the corresponding data is available for use within the procedure, with some of the data possibly already being read at an earlier time. box 375 ("cmp LDF and WSD ") corresponds to the execution of the calculation of LDF and WSD based on the procedure described above The input data for this calculation procedure includes the data for unallocated wheelsets taking into account, if applicable, the data of at least the last two wheelsets of the previous vehicle. The Applicant prefers that in view of the above, appropriate uncertainty values be assigned to the wheel sensor data and, more particularly, that adequate uncertainty be assigned to the wheelset data of the previous vehicle because of the clearance of the buffering.

The branching of box 376 ("BPD x?") Depends on the existence of the buffer profile data (here "BPD"). If the BPD are available, they will be at box 377 ("proc BPD") is read and processed to identify at least one end buffer interface and assign it a longitudinal distance from the corresponding wheel group of the vehicle being analyzed in this identification. The analysis of the BPD uses the box 375 calculated LDF and extends to a data series that corresponds to a value that exceeds the maximum vehicle length by a predetermined margin. It is of course possible that no buffer interface will be found due to measurement disturbances or, more frequently, that this will not exist, as in the case of articulated trucks on articulated bogies or in certain passenger vehicles.

A vehicle model suitable as a "candidate" (here "CVM" for candidate vehicle model) is a vehicle model which, if possible, corresponds to a vehicle whose model has not yet been identified. A list of these vehicle models (here "CVML" for Candidate Vehicles Models List) is available at Box 378 ("cre / upd CVML") created when the box 374 did not enter after the box 373 was entered the last time. The other way around, if the box 378 after an entry in box 374 will enter after the box 373 was last entered, then the list is updated as explained below. Candidate vehicles are added to the list based on the wheelset distances WSD and the buffer positions, if these are in Box 372 or box 377 were made available. The CVMs are selected according to the matching wheelset spacing WSD and the buffer distances, if applicable, taking into account the uncertainties of these values. The search for vehicles that are compatible with the corresponding WSD and, if possible, with the buffer distances is performed using a data structure called "CVMSD" (for "Candidate Vehicle Models Selection Dataset)." The content of CVMSD is one Subgroup of the content of the vehicle database and can be organized in many different ways to effectively carry out the creation of CVML or an update by using the appropriate efficient algorithm for vehicles without a symmetry of wheelset positions and, if applicable to their buffer distances, with respect to the half-length plane, one can have two data series in the CVMSD, each corresponding to the different scanning directions, or just one series of data, and then the data is searched from both ends of the vehicle, a matching value at that point associated with each member of CVML, ie ba based on the sum of the squared differences between actual wheelset offset values and the CVMSD values. Such quantifying match values may be used in further steps of the VI application to sort the candidate models. In the exceptional case that no candidate model is found during the creation of the CVML on the basis of the first two unassigned wheelsets, it is advisable to consider only the first wheel group and the CVML remains empty.

The branching of the box 379 ("CV w + WS?") To box 374 ("Get WTD") occurs when at least one candidate in the CVML exists with more wheelsets than those whose data has already been considered. The branch to box 374 would not be done if, for example, the interrogated vehicle is a long semi-trailer on a double-axle articulated bogie, or if the vehicle is a long double-axle rail vehicle. The box 374 indicates the reading of other unassigned wheelset transit data, as in Box 373 , and this additional data will be sent to Box 375 used in another calculation of LDF and WSD, which correspond to a higher wheelset number. The reason for this is that LDF and WSD are usually more accurate for a four-axle vehicle when all wheelsets are considered, and not just for dual-axle vehicles. After returning to box 375 from box 374 in the current vehicle identification process, the processes of Box 377 or they can be re-run for the LDF results for multiple wheelsets. The process of M after the addition of further wheelsets represents an update of the CVL in which the previous conflicting candidates are sorted out by newly calculated WSDs according to the wheelset number.

box 380 ("get CVM IDS") corresponds to loading from the vehicle database, or from a reduced specially created vehicle database, the complete identification records (here "IDS" for Identification Data Set) of the vehicles that are in the current CVML. The identification records may be arranged in various ways and may generally be defined as a collection of specific vehicle features that may be used by the VI applications to select the correct vehicle identification method dells from the candidate list can be used. The required information contained in the IDS depends on which optical sensors are installed in the system and which algorithms have been used in the vehicle identification process. It's also possible the ISD is not at box 380 or retrieve only a portion of the IDS and then retrieve the IDS information and data at various later times in the vehicle identification process. If Radprofile data is available, the box points 382 ("WPD x?") To box 383 ("get & proc WPD"), which corresponds to the processing of wheel profile data to produce approximate wheel diameters and wheel profiles that must match the wheel data of the current vehicle candidates.

If axle load data is available from the appropriate hardware and software, as discussed below, then the box references 385 ("XLL x?") To box 386 ("get XLL") where the load data is retrieved. The load data of the axles serve in principle to differentiate between vehicles such as electric locomotives whose mass and mass distribution on the axles is almost constant. In addition, one can remove from the CVML any vehicle model whose weight, taking into account the actual weighing accuracy, is clearly below the minimum weight of a candidate vehicle with no load or optional equipment.

The box 384 ("upd CVML") refers to the updating of the CVML by eliminating candidates that do not meet the compliance criteria based on wheel diameter, wheel body profiles or axle loads within given tolerances.

The branching of box 381 ("CVM n") is based on the number of members of the current candidate list of vehicle models. If at least one candidate model exists, the box points 381 to box 387 while she's boxing elsewhere 395 points. Around 9 not to be too complex, a box was like 384 when leaving the box 383 omitted and the same branch of box 381 was not specified after each process, as well as in 378 creating or modifying a CVML.

The characteristics of hardware and software that take over and manage vehicle images are addressed later in the document, with the discussion of imaging data being limited to their use in the vehicle identification procedure. The box 387 ("build IMA") corresponds to the preparation of images intended to contain the distinctive alphanumeric markings corresponding to UIC Code Leaflets [057, 058, 059] for passenger, freight and towing vehicles.

Although an OCR (Optical Character Recognition) function can be applied to all images of the passing rail vehicle, the limitation of such a process to areas where the presence of such a mark is expected makes the OCR and subsequent recognition processes more efficient. Thus, to define a restricted area at the entrance to the OCR process, the search coordinates of the marker areas (here "MSA" for Marking Searching Areas) from the IDS are used for the candidate vehicles in the current CVML, and it is advantageous to avoid multiple OCRs Processing of some image areas in relation to different candidate vehicles, the MSA for all current candidate vehicles overlap such that for each vehicle side the coordinates of one or more areas are generated (here IMA for images of possible marking areas) for which the OCR MSA and IMA shapes are preferably rectangular to facilitate some aspects of the mark detection process 375 Calculated LDF and calibration data of the imaging devices are used to determine the appropriate frame times and pixel ranges that must be retrieved for the preparation of IMA. The IMA preparation may include retesting the fetched pixel data array to provide the OCR application with screen arrays whose horizontal and vertical pixel lines correspond to certain predetermined pixel values in millimeters measured at the marker area, assuming a certain approximate distance of the pixels Vehicle surface of these imaging devices. Alternatively, the horizontal and vertical scale factors should be calculated by IMA and provided to the OCR application. In both approaches, OCR is able to use and generate approximate absolute dimension values for characters and their positions.

box 388 ("OCR IMA") corresponds to the OCR processing of IMA. To date, several different OCR methods and algorithms have been developed and are not discussed here because they are known to those skilled in the art and because original OCR engineering development is not considered necessary for the present application. The choice of a suitable OCR technique should take into account that the characters that are to be recognized may have a low contrast to the background, that their font is not strictly standardized, and that their size and spacing from each other are minimum that a considerable freedom of choice is given by the UIC code when marking rail vehicles. In addition, it is important to note that the icon images that are to be detected may have been altered by wear, aging, and dirt, and that the images may have a number of features that may interfere with the intended recognition process. The UIC code does not allow a few other formats for marking rail vehicles and therefore the number is limited to the resulting symbol pattern, which can be taken into account within the OCR procedure. Full and accurate recognition of all identification symbols is not necessarily required, and the applicant prefers that the OCR process has an output that only specifies a recognized symbol when the detection uncertainty is very low and when the symbol is less likely in less certain cases contains matching symbols that should be given as possible with a possible probability estimate. The coordinates of the symbol should also be included in the output so that the information that the subsequent processes use in the vehicle identification process is as rich as possible. The choice and customization of the OCR application typically requires a series of trials and adjustments made with a series of sample images as determined by the system implementation in the current manner. The appropriate level of "learning" of the current OCE application must also be defined as part of the development and / or customization of the OCR software module.

The box 389 ("proc OCRO") corresponds to the processing of the OCR output data (here OCRO for Optical Character Recognition process output) in combination with the IDS information to achieve a very high success rate in the vehicle combination, as well as a very low frequency of false identifications and a relatively short processing time. This module can be designed in many different widths and above all can use different combinations of methods, algorithms and heuristic techniques. The Applicant therefore prefers to discuss the process of Box 389 to restrict to a series of remarks and incomplete possible techniques or parts of algorithms, and to leave the design of the module to the appropriate engineers. A first note about the software application of 389 is that only the members of the CVML may be considered candidates for the match of recognized UIC marker symbols. A second note is that the main objective of the whole vehicle identification process is the detection of the rail vehicle model, while the other information occurring in the UIC marking of rail vehicles, such as the operator of the rail vehicle or the unique serial number, thanks to which a particular part in a fleet of the can be useful for some of the added value functions resulting from the integration, but are not necessarily required to achieve the main objectives of the system. A third note is that the UIC mark must be present on both sides of the rail vehicle, and that imperfect recognition of part of a UIC mark on one side can be supplemented by the OCR results of the other side, even if also this second mark is incomplete. One of the possible approaches that the Applicant has for the process of Box 389 is based on a sequential consideration of the members of the CVML. One or more symbol patterns for a particular candidate vehicle may be found in the IDS as pre-loaded, ready-to-use information, or may be extracted from the IDS information along with the applicable rules that may be readily extracted from the UIC tag leaflets , Then you can search for a match of the OCRO data of both sides of the vehicle with such patterns. The identification of additional variable symbols for complete recognition of the UIC code may then be done, further based on the OCRO data for both sides, the simplifications being provided by the already allocated non-variable symbols and by the combination of their positions with the limited placement possibilities the variable symbols according to the possible patterns can be used. The control icon of the UIC tag may be used to validate the previous icons and sort out certain combinations of symbols that may result from the detection uncertainty of one or more characters. By using some standard Artificial Intelligence techniques, such a process can be organized for each vehicle model candidate as a heuristic search in combinatorial problem space, where the heuristic rules may affect the preliminary organization of the OCRO data. Fuzzy logic and probability estimates can also be used in exploring the problem space. Applying a relatively fast and crude solution finding process for OCRO compliance with each candidate identity prior to performing sophisticated or time-consuming alternatives may be a good choice, since the likelihood of the task being very complex because of a very unfavorable OCRO is relatively low. An example of how to make the above process very fast while retaining most of its efficiency is to limit the initial search for non-variable character sequences to near-horizontal sequences, all or all but one of the non-variable in correct order Contain signs for each vehicle page while leaving boundaries for the characters size and the character spacing sets. Applicant generally suggests finding the solution-finding capabilities for the application of Box 389 so that the corresponding average and maximum processing time is shortened, and thus does not slow down the primary vehicle identification procedure, which is designed to produce as quickly as possible an output that is a vital input to the other system processes. In this approach, the incidence of vehicles not identified in this first stage will increase, but this would only subject the number of cases to the second VI application, which will be discussed below. Of course, the criteria for determining the sufficient safety of a vehicle identification are for the process of Box 389 and the Applicant proposes to strictly observe them, as the definition of an unidentified vehicle at the primary VI level is clearly preferable to misidentification.

The main results of the primary VI procedure, i. LDF series, the assignment of wheelsets to Vehicles and the identifications of vehicle models, can be found in a number of different data structures are arranged their design has some simple requirements and desirable features should be fair, such as the simplicity of use in various applicable system processes while the Data collection or the VI process are still in progress, as well as the compatibility with the functions of a secondary VI procedure involving more than one vehicle from a group of together lying wheelsets can define those in the primary Identification procedure associated with an unidentified vehicle were. A possible Way to organize these results during their sequential generation is a list in which each element of a wheelset in the order the discovery while corresponds to the trace sampling and contains scoreboards that reference other data structures refer to information about vehicles that identified or can not be identified, directly or indirectly accessible is. It can also be handy to use the list of wheelsets to indicate a value indicating whether the corresponding vehicle was identified or not.

If the process of box 389 If a secure vehicle identification is reached, the branch of Box opens 390 ("VI OK?") To box 391 (flag IV), ie the identified vehicle is flagged and the corresponding wheel group is assigned to this IV and the corresponding data is written to the vehicle data structure or structures, if applicable, as described above. Unidentified vehicles eventually correspond to a wheelset series for which the primary VI procedure could not assign a secure identifier. A new unidentified vehicle is "created" as a result of a failed assignment of a wheel group to an identified vehicle, provided that the previous vehicle has been identified, or in the special case of a start of the train all or some of the unassigned wheelsets that could not be assigned to an identified vehicle were assigned to the previous unidentified vehicle 395 ("PV to UV") is boxed by 381 reached when the CVML is empty, or from box 390 when the process of box 389 could not achieve a safe vehicle model identification. If the previous vehicle was not unidentified, the branching will result from Box 395 to box 396 ("cre UV"), where the creation of a new unidentified vehicle takes place before continuing to Box 397 ("add WS to UV"). Alternatively, if the previous vehicle was an unidentified vehicle, the branching will result from Box 395 directly to box 397 where all or some of the wheelsets that could not be assigned to an identified vehicle are assigned to the corresponding unidentified vehicle.

It is better if the actual number of box 397 Wheel Sets to Assign a Safe Minimum Number is because if the unassigned wheels are an unidentified vehicle at Box 397 are associated with one another, do not belong to the same actual vehicle but partly belong to a leading vehicle and partly to a final vehicle, the tracking of one of these two vehicles is not identified at this point. Therefore, the number of wheels associated with a previous unidentified vehicle may be limited to one or a number that may be defined as a function of the contemplated WSD, with the rules based on the combination limits for wheelsets in the existing rail vehicle population are based.

When leaving the box 397 leads the junction of box 398 ("WS left?") To the box 371 , if further unassigned wheelsets have to be considered. If, on the other hand, there are no unassigned wheelsets, the current unidentified vehicle becomes Box 399 ("UV flag") and the primary VI procedure ends at Box 370 ("END").

When leaving the box 391 leads the junction of box 392 ("is PV to UV?") to box 393 ("UV flag"), if the previous vehicle was an unidentified vehicle, as no such wheels are assigned to such a vehicle. If, on the other hand, the previous vehicle has been identified, Box leads 392 to the Branching of box 394 ("WSD left?"), Which also when leaving box 393 is reached. If more wheel sets have to be assigned, the branch will lead to Box 394 to box 371 ; otherwise, the primary VI procedure ends with Box 370 ,

professionals It is clear that the primary VI procedure described above is different Ways can be formulated or varied more or less, e.g. by doing another detailed procedure for creating the candidate vehicle list used. In particular, it is also possible to create rules the search for CVML elements by the detection of certain Facilitate patterns in unassigned wheelset distances. The applicant would like to emphasize that the primary VI procedure fast and reliable should be the time for the identification of identified vehicles to a minimum to reduce the other system applications described below their processes for the individual vehicles whose model has been identified begin can.

Out the one in the introduction of the document, the VI procedure described above does not use any identification technique, requires the markings, and no information systems, today only for some rail vehicles or trains or railway infrastructures available are. It is possible, however, and can be handy, such additional sources of information to integrate into the system. For example, if radio frequency markings passing in a significant part of the system installation Rail vehicle can be installed, so or the corresponding Readers be installed and the identification data can easily be changed Version of the primary VI procedure used, however, in principle for the vehicles that are not are marked, remains the same. It is also possible that the system train composition from the train itself or through an information system gets that has this information on the train composition itself or retrieves it elsewhere. In any case holds it is the applicant for recommended and useful, that the recognition processes described in this document at least partly also for Rail vehicles performed be that identification by markings or others Information systems is given as a mismatch with a very reliable Measurement like the wheelset distances accidents or malicious Can reveal facts. So it is possible, for example recognize discrepancies in train composition this way which are based on a problem at the yards, or even forgeries the vehicle's marking, its marking and record. Of course in such make a false identification creates a message to the corresponding remote control centers or information systems would.

5.5 Secondary vehicle identification procedure

The by box 235 in 3 The secondary vehicle identification procedure that has been initiated serves to attempt to identify the vehicles identified as unidentified by the primary vehicle identification procedure. The secondary VI procedure can be more successful in identifying, because it has more time to solve complex cases, and because additional measurements and additional information from the vehicle database can be used, which were not taken into account in the processes of the primary VI procedure. In the preferred implementation of the method, no identification is attempted by the secondary VI procedure for an unidentified vehicle if the corresponding CVML exits from the CVML build or during the update process as in Box 378 or 384 from 9 was empty. These vehicles are referred to herein as "unknown vehicles" because no applicable candidate model was found in the group of vehicle models known to the system through its vehicle file, based on the fundamental information about them, ie, wheel spacing and possibly buffer positions, wheel characteristics and axle load Reason why a vehicle may fall into the category of unidentified rail vehicles is the lack of relevant information in the vehicle database and the presence of one or more faults for the corresponding vehicle model in the vehicle database The applicant prefers to use the secondary VI method It is also better to use two identifiers if you want to use sorting identities that do not conform to the applicable identifiers be defined for the secondary VI procedure. The first level of identification corresponds to the selection of only one candidate for the vehicle database and the correspondence of OCRO data from both sides of the vehicle, so that all markings related to the model are identified. The second and lower level of identification corresponds to the selection of a candidate only without disagreements with applicable recognition features, but without a positive match of all mark symbols relating to the model. The identification stage may be used in some procedures of the method described below. Identification data by electronic identification or Infor Train composition information from external systems may be used at this point to pass the vehicle identification from the first to the second stage if the OCRO data does not produce a result.

Certain geometric features can in the secondary VI procedure for sorting out candidate identities and to increase the validation security for one or more candidates are used. The for this Purpose to use geometric features are stored in the vehicle database and can for the most part correspond to the characteristics used later in the document, to the position and orientation over time of the car body to determine. The presence of these features for that too Valuable vehicle requires that the vehicle data from the sensors the image forming apparatus and retrieved the three-dimensional gauges and with the LDF be processed, as explained below. In case of a mismatch the only identity a vehicle that could be assigned and an applicable one Identification feature leads for classifying the vehicle as unknown, i. like that Vehicles, for in the vehicle database no candidates were found, see above.

5.6 VIS or NIR imaging of vehicles

As already explained above and also discussed below in the document are pictures of rail vehicles in visible (VIS) or near-infrared (NIR) bundles of the electromagnetic Radiation spectrum in the process for various fundamental or optional purposes, including for reading the unmistakable UIC vehicle markings, for the Reading other markings of combined transport vehicles, tractors or containers, for the reading of signs with codes for dangerous goods transports, for the clear Recognition of vehicle characteristics, for the determination of the driving curve from car bodies and for the Recording vehicle images sent to remote control centers and / or other information technology systems can be sent. The Subsystem for the determination of VIS and / or NIR images of vehicles must be for recording be suitable for images whose resolution and contrast for the o.g. Purposes under all applicable working conditions and especially in all weather conditions, in all natural light conditions and for the total train speed range specified for the system operations is, is suitable.

The Positioning of the picture equipment for all planned system applications and compliance with safety regulations for the different lighting of passenger trains and train crews are two more important aspects to consider when choosing the image sensors and the lighting equipment Need to become.

The most stringent requirements for digital image resolution are provided by the use of images as input to the OCR process. However, the actual choice of image hardware is not simply based on the smallest image character size, expressed in pixel units, but it depends on the actual MTD (Modulation Transfer Function) of the imaging device, the properties of the OCR algorithms, and the characteristics of the characters to be recognized from. Although the suitability of the imaging and lighting equipment can be quickly assessed by those skilled in the art based on their knowledge and experience, the Applicant recommends choosing the imaging and lighting equipment after their field test under the actual working conditions and considering a large number of rail vehicles. However, some sub-information and discussion will be provided below on the VIS or NIR vehicle image subsystems. Line cameras, that is, imaging devices that use a line of photosensors instead of the two-dimensional array of photosensors, are often used in automatic vision applications where the image target performs smooth motion in front of the sensor, as in the case of articles on a conveyor belt. Linear imaging equipment has been used successfully for road and rail vehicle registration [065, 066] and is also very attractive for implementing the system. One of the most obvious advantages in this case is the creation of a single continuous image along a direction of motion rather than a series of two-dimensional images that are complex and can not be inserted into a composite image. Therefore, there is no need to synchronize the image capture with the presence of certain objects in the field of view of the imaging device synchronously. Long objects that extend beyond the direction of movement can be imaged without discontinuities. Other benefits include reducing the area to be illuminated, the potential for large transmission rate savings and the overall size of unprocessed image data, as well as easier protection of the exposed parts of the device (eg the lens) from atmospheric and direct sunlight. The above advantages are also the reason that the applicant, the linear Bilderzeu Although the system may also be implemented using imaging imaging cameras for areas, it is preferred that the best practice of this invention be discussed below.

10a and 10b show a vertical and a horizontal section of the passing vehicle passing through a set of ten line cameras 440 . 441 . 442 . 443 . 444 . 445 . 446 . 447 . 449 and 450 is recorded. The plane for the linear image generation, ie the locus of the points which are imaged at different distances from the camera, is in the example of FIG 10 vertically and the imaging devices have a zero slope in the horizontal plane (the term "imaging plane" is used here, even if, geometrically speaking, this locus does not lie in a plane, but has a finite strength and therefore more resembles a blade). In practice, these two conditions are not required for system implementation, although it is advisable to fine-tune them and to compensate for a small deviation in the corresponding angles during calibration, as explained later in the document. The line cameras positioned for recording the sides of the rail vehicle, as well as those with the numbers 440 to 447 in 10a The actual number of lateral cameras and their positions, their fields of view and the resolution are not strictly defined in this text and the applicant prefers to make some technical remarks here that support the actual detailed design of the system implementation.

As mentioned above in the general description of the invention and the desired features of the system, the devices installed for the SMI should not limit the possible locations for the SMI to a short portion of the rail length. Most importantly, the system should be compatible with dual or multiple railroad tracks, and the imaging devices should be positioned on either side of the appropriate rail track so that the UIC markings on both sides of the vehicle can be read. For most European lines [062], for which the installation of the system could be interesting, has the separation 462 between the centers 451 and 452 the rails have a minimum value of 4000 mm, which means that at one width 471 the boundary line profile of the infrastructure according to the UIC standards for kinematic boundary lines [050, 051, 053], the gap distance 459 between the page boundaries 455 and 456 for positioning a foreign object is almost 500 mm.

The lateral position (distance when passing from 451 from the vertical plane of symmetry of the rail) of the vehicle markings agrees approximately with the side of the execution model of the vehicle body (as in the example profile 453 ) that just within the corresponding gauge line profile 454 lies. The actual length of the imaged length on the vehicle side depends on the distance 458 , from the field of vision 448 and from the angle 464 in 10b between a parallel 465 to the rails and the imaging plane 469 the line camera 460 , one of the two lateral cameras 444 - 447 in 10a , The pixel number of the line cameras suitable for this application should be 1024, 2048, 4096 and even higher, giving them a target spacing of the pixels of about 2, 1, 0.5 mm or less with a camera distance of 1 meter and a field of view 448 of 90 degrees. Consequently, in consideration of the features of the characters to be imaged for the OCR process, the minimum number of lateral cameras for imaging the entire vehicle side is not limited by the resolution of the line cameras, but depends mainly on the maximum desired angle values 448 and 469 as well as the limited distance of the camera. If one assumes that the distance 457 between the back of a lateral camera and the positioning boundary 456 is practically zero, then it is advisable to select the lateral cameras and their housings with a short distance between the imaging sensor and the camera back to increase the distance. The angles 448 and 469 determine the smallest angle between the line of sight of the camera pixels and the vehicle side surface. These last angles are the minimum angles for extreme pixel values and should be limited to avoid possible distortions and quality losses of the mark images. It must also be considered that at an angle 469 significantly smaller than 90 degrees, the images do not contain some vehicle details that could have been hidden by relative abrupt changes in the vehicle profile with respect to its longitudinal coordinate, unless the number of cameras is doubled and half of them look in one viewing angle in the direction of movement and the other half in the opposite direction.

The minimum required overlap of the imaged lines with adjacent cameras (eg 446 and 447 ) at the shortest installation distance is related to some features of the software that processes the corresponding data, and especially to the recognition of the markers that are at the edge of the field of view of the line cameras. One option is to leave the overlap large enough to ensure that a marker line of the characters or an entire marker is made is more than one line, completely within the field of view of at least one of the two adjacent cameras. Alternatively, one can reduce the guaranteed minimum overlap to less than the size of the corresponding marker and assemble the relevant juxtaposed images before attempting to recognize the characters. The coincidence of two adjacent images obviously depends on the imaged area position, and superimposing images usually requires grading, which is facilitated by assuming the approximate location of the imaged area of the vehicle database or three-dimensional measurements, such as will be explained below. The corresponding software algorithms that can be implemented to manage the composition of juxtaposed images are not discussed here because they are well known to those skilled in the art of processing multiple digital images.

Of course you have to in the implementation of the system a choice between color images, NIR images, b / w (black and white) images and some special variations of these are taken as these Choice of both the imaging devices and the lighting devices and focus on the balance of cost, performance, reliability, Availability and ease of maintenance. The marking of rail vehicles was not up to date in terms of their characters and background colors standardized and also has the consequences of character recognition through the computer vision not considered. In this regard, color images are beneficial as they enhance the OCR process in the worst cases in which the characters and background colors for the VIS or NIR imaging with very poor contrast appear in a gray scale. Another advantage of using color cameras is recognition the distinctive background color of marker signs of dangerous goods. The high speed imaging with color lines however, it is more expensive than B / W and needs more of image data and generally needs stronger lighting. The usage suitable color filter can in the above worst cases the Improve contrast of images taken with B / W line cameras. The combination of two black and white cameras with different spectral Reaction, i. using different filters and a single one Model of a line camera, so can represent a solution, even if she higher Cost and a bunch of extra Difficulty in data processing compared to use from simple black and white cameras. As the applicant concludes empirical Experiences with statistically representative vehicle records missing, beats He suggests that the choice of the actual Type and model of the camera together with one of the lighting units is taken only after a few preliminary field tests. artificial Lighting is at least then for the system needed if the natural ones lighting conditions are not sufficient. The lighting intensity at the destination is a crucial one Factor of image quality. A more Lighting will generally make possible the signal-to-noise ratio of the To reduce image pixels and thus the contrast resolution too increase, and possibly so much so that images with a gray scale in practice sufficient are good to opposite Consist of color images. Furthermore the use of intense artificial lighting reduces the contribution of natural Light relative, which leads to the advantage of rapid adaptation the camera sensitivity is no longer necessary.

And after all allows intensive lighting will reduce the opening of the Iris, resulting in an extension of the depth of field and an improvement the picture sharpness leads.

Applicant wishes to emphasize that the illumination required for this application is much greater than for ordinary imaging applications of similar geometry because of vehicle motion. The electronic shutter speed is obviously shorter than the time interval between the triggering of two line imaging devices, which is inversely proportional to the vehicle speed needed for the same image resolution. For example, imaging at 1 meter distance with a 2048 pixel line camera with a 90 degree field of view implies a vertical pixel pitch on the target area of about 1 mm, with the same resolution in the direction of motion at a line sampling frequency of about 22 kHz at 80 km / h or33 kHz at 120 km / h. Requires. Thus, taking into account the various geometries, train speed, and resolution above, the electronic shutter speed of this image application is usually in the range of one-hundredth of a thousandth of a second, which is in fact a fairly short exposure time. However, it should be emphasized that the requirement of intense illumination due to vehicle speed is not associated with the use of line cameras, since two-dimensional imaging would also require a very short exposure time (due to fast shutter speed and / or an impulse light source) Blur the images in the longitudinal direction to avoid. The points 461 and 463 in 10b (not shown in 10a ) represent two vertical series of lighting devices mounted practically on the side of the series of side cameras, as well as point 460 , Their distributions of light, hinted at by the Lobi 466 and 467 are shown, converge towards the image target area and thus provide a uniform and evenly distributed lighting. In general, the use of multiple illumination sources and linear conti Continuous or near-continuous sources of illumination to achieve uniform and intense illumination while minimizing the disruption of train passengers and crews. The use of pulsed LED lighting devices should be considered, especially when using NIR cameras, as they offer an advantage in terms of low power consumption, high availability and long service intervals. In general, the illumination sources of the lateral cameras could alternatively be positioned between the cameras, rather than laterally of them, or they could be arranged in more than two vertical rows.

The choice of the optics and the design of the lighting system should take into account the possibility that the imaging process will be in the direction of the axis 451 extends the vehicle, at least in the image position of the marking labels for dangerous goods transports. The vertical position of the lower side imaging device may just barely be enough to detect the lowermost markings of a vehicle, or it may be as in the case of 10a in that the underside of the wheels and part of the track is included outside the railway tracks. The advantages of low positioning of the lowermost lateral cameras are related to the use of the corresponding image data for calibration, as well as to the monitoring purpose of the integrity of the system, as discussed below. VIS or NIR imaging of wheels and bogie elements may be useful in data processing in the detection of axle-related overheating, as discussed below.

The requirements for selecting and positioning cameras for picking up the vehicles from above, here called "bird's-eye view camera", are different from those discussed above for the lateral cameras, and in fact, the images taken by these cameras do not become character recognition Their installation is principally useful in the context of determining the vehicle body movement and recording documentation of passing vehicles, particularly those vehicles for which a boundary profile alarm has been triggered.The resolution of the bird's-eye view cameras must be evaluated taking into account the maximum recording distance, Case of observation of a roof of a flat car and with a positioning of the camera similar to that in 10a for the cameras 449 and 450 is shown at a distance of about 5 meters. The place of the cameras 449 and 450 in 10a is not an exact indication, but an acceptable approximation to a suitable solution that is also compatible with the GC gauge line profiles according to UIC standards [053], where the top of the datum line profile is significantly higher and wider than the gauge line profile 454 That's about the same 10a is drawn. Such an approximate position is also recommended to prevent too close proximity to the trainline, given safety and maintenance considerations.

A certain distance 470 in 10b between the positions along the rail of the two camera groups, which are positioned on the opposite sides of the rails, can help to avoid the cameras on the one hand, light from the lighting devices on the other side at a small angle to their linear recording plane falls. The distance 470 can, for practical reasons, be widened to a few meters if the system software uses acquisition features to evaluate the trajectory (position and orientation) of the vehicle body.

The Processing of raw data collected by line cameras requires a precise assessment the vehicle movement, in this case at least by the above described LDF is delivered. The synchronization of the data transfer is addressed below in the context of the data transfer electronics.

High resolution fast Line cameras are offered by different manufacturers. Some examples include 1024 and 2048 pixel versions and linear sampling rates of up to about 50 kHz, for example, in the series Piranha CL-P1 [955] from DALSA Corporation of Waterloo, Canada.

DALSA Corporation also manufactures ultra-sensitive line cameras, such as the Eclipse EC-11 [956] series and the DALSA HS-41 [957] based on Time Delay and Integration (TDI) technology. These cameras are particularly interesting for the implementation of the system because they require lower illumination intensity, but they must be synchronized with the vehicle speed in a range of 2-4% to prevent deterioration of image quality. Such synchronization can be achieved in different ways within the system, and more particularly by being inserted into the data acquisition module for wheel sensors of a real-time estimation of vehicle speed by a simple and fast algorithm that is not as accurate as the method described in the LDF calculation , but sufficiently accurate for the synchronization of TDI line cameras. Such an algorithm For example, it may calculate the current average transit time transit time of any wheel set between two adjacent wheel sensor pairs.

5.7 Three-dimensional measurements the vehicle geometry

5.7.1 General remarks to the three-dimensional vehicle measurements

The Method for determining the boundary line profile described below a rail vehicle requires performing three-dimensional Measurements of the vehicle geometry and especially of the car body. More accurate said, the method requires the generation of a data series, here "3DD" for three-dimensional Data called the coordinates of a vehicle surface point in an infrastructure-based three-dimensional coordinate system and the corresponding time. Given the use This data generally does not require the whole Vehicle area too measured. The appropriate measurement system should have the 3DD at least for those Create vehicle parts that transit in the space between two spatial wrapping lines lying at a certain distance inside and outside of Reference boundary line profile according to UIC [050, 051, 052, 053]. The applicant makes no binding specifications for these distances, the define the three-dimensional measuring range, since there is a certain Advantage of keeping these larger when according to UIC codes is required to determine with the minimum goal whether a vehicle has a part outside the corresponding one Boundary line profile. The main advantage of the extension of the 3DD measuring range towards the vehicle is the availability of the 3DD for a larger number of vehicle features used in the process described below for determining the trajectory of a vehicle body can be used. On the other side allows the extension of the 3DD measuring range via the Reference profile, the actual Length too Measure to a mechanical part in an improper way on the Perimeter profile stands out rather than just finding it on impermissible Way stands out.

in the Principle should not be a vehicle feature (including the load) the applicable Profile boundaries hurt, but given the fact that the System has the task of risk reduction and that complexity with the Slimming the size of recognizable above parts, it seems right that a practical Goal regarding the size of a to be recognized part.

The Probability of having a particular object in a given Position is not detected depends precisely from the "density of space the measurements ", from the measurement of the intersection area, the object size and the object orientation opposite the 3DD meter from. The required recognizability of objects can be combined the 3DD measurement performance or by an empirical performance criterion expressed be that certain supernumerary Specifies default structures used in the bump test of the system can be. Given the differences between alternating feasible 3DD measurement systems and the current lack of a standard for detection of protruding objects, If the applicant prefers to provide some alternatives for the implementation of Discuss 3DD measurements and leave it to him Design of the system implementation, the best alternative to choose, too dependent from the definition of the user's accepted performance requirements.

An important feature of the measuring means for the 3DD measurements is the measurement uncertainty, which in this case is not just a single value, since it is usually different for different directions. For most three-dimensional measurement systems, the measurement uncertainty is defined along three axes whose orientation is determined by the position of the instrument relative to the element to be measured. Another special point to consider is that the measurement is typically done on a finite size point on the measured area, which point is part of a circular or elliptical optical beam in most measuring means. The orientation of the measured area and its possible curvature affect the measurement in different ways, depending on the measurement system used. In addition, the 3DD uncertainties are otherwise affected in different measurement systems by the presence of particles in the atmosphere. Another instrument-specific uncertainty source is associated with measurements taken at the edge of an object when another object is in the background and affects the outcome of the measurement process. Finally, the distance between the instrument and the measurement target also influences the measurement uncertainty. It may therefore be necessary for the inherited data, along with the calibration and data configuration when used in 3DD processing, to also allow uncertain numbers, which generally depend on the 3DD coordinate values. Another key feature of the alternating 3DD measurement systems is the measurement time or, more specifically, the time at which the measurement object, generally by an optical detector, is sensed. The measuring time must be compatible with the maximum vehicle speed to obtain the desired resolution and to avoid unnatural measurements due to the displacement of the measured object during sensing. Although the Applicant does not consider it appropriate to set strict specification specifications for the resolution and sensitivity of the 3DD measurement system, it is worth noting that the UIC Leaflet for Leaflet Profiles and the explanations below relate to the 3DD measurements in that the geometric 3DD accuracy in the direction perpendicular to the boundary line profile is much more critical than the geometric accuracy in the direction of movement or tangential to the boundary line profile. Applicant believes that a reasonable 2σ accuracy range in the 3DD measurement for the component orthogonal to the perimeter profile can be between 5 and 15 mm, depending on the desired limits for missed detections or false alarms, since these thresholds do not extend the applicability of this invention limit.

5.7.2 3D measurements, the based on stereoscopic images

A well-known method for the determination of three-dimensional geometry measurements is to use a stereoscopic vision, i. the illustration of the measurement target by two or more cameras and the reconstruction the surface geometry by finding the three-dimensional location of a feature that coincides with his two-dimensional position for the images determined by said two or more cameras match. In the special application could the vehicle image are well performed by linear imaging devices, as an alternative to the more often used area imaging devices. Lots Algorithms are known and described in publications, to solve the opposite problem, which makes up this measurement method. However, the applicant is in general against the use of stereoscopic vision within the method to identify structures that are dangerous about vehicles can protrude since the possibility 3DD measurements successful for the critical feature regarding the gauge line profile perform, depends on whether this feature is mapped correctly, as a function of its Shape, its optical surface properties, the illumination and position of the imaging tool. Stereoscopic three-dimensional measurements at specific positions on a vehicle, however, as an option in the below described method used to determine the position and the orientation over time of the vehicle body and axle-related Parts is used.

5.7.3 Image of light patterns

A Another widely used method for determining 3DD measurements is the image through one or more cameras of a light pattern, which is projected onto the measurement target. A special configuration of this Type of measuring arrangement has been described [067] for the purpose of recognition of rail vehicle structures that protrude beyond a boundary profile. Such an arrangement, based on a orthogonal to the rails Line lighting and picture by a camera showing the rails can see along, can for the system may be inappropriate, at least because of insufficient resolution in the longitudinal direction. The applicant concludes the use of this particular arrangement of 3D imaging method not but he does not prefer them because of the lack of reliability in terms of geometric variability of the relevant Measurement objectives. additionally this requires the imaging of light patterns a very intense structured lighting when used for fast-moving targets, in direct sunlight and with a great variety the optical properties of the measured surfaces is, and is therefore hardly related to the security requirements to arrange with the eyesight of train travelers and crews.

5.7.4 Optical series lists

Series of optical barriers, such as a comb-like series of photoelectric sensors, are often used in automation systems for quickly detecting the position of an object, but they can not be easily used in the present application because they do not provide individual information about the position of the detected feature along one Deliver interrupted light beam. In principle, however, one can install series groups of optical barriers with different angles of their optical beams and the vertical rail axis, thus implementing some sort of optical tomography. The advantage of using multiple optical barriers is the relative simplicity as well as the low cost of the individual devices, but Applicant does not care to use this measurement solution because of the complexity due to the large number of rack series that would be required, and due to the resulting considerations regarding availability and maintenance, also with respect to a possible Settling various types of dirt, dust or dull grease on the transmitters or the receiving surface.

5.7.5 barriers with fixed Laser distance sensors

There are several sensors that work with lasers and seem capable of achieving fast repetition of the distance along the laser beam path. One of the widely used families of laser distance measuring devices ("LDM") that can be proposed for this application is based on a time measurement by a laser pulse that reaches the target and then returns to the instrument after it diffuses through the target surface was reflected. These measuring instruments, often referred to as "time-of-flight" laser rangefinders, are offered by different manufacturers in different models, meeting various requirements for maximum and minimum measured distance, laser capacitance, repetition rate, laser spot size and accuracy. For example, many models of high-speed laser distance measuring devices are available from Riegl Laser Measurement Systems of Horn, Austria, such as the model LD90-3100VHS-FLP (measuring rate of up to 2.0 10 ³ sec ^{. 1} ) [958] and the model LD90-3100EHS-FLP (measuring rate of up to 1.2 10 ⁴ s ^-1 ) [959]. These two instruments are actually suitable for much longer than the ranges required in this application, but they have some important features such as an input for external control, single-jaw measurement, and the last-impulse option that they discussed for those in this area An externally triggered single-pulse operation allows each measurement to be timed precisely (see discussion in Section 5.18 below) and prevents any kind of unnatural values resulting from vehicle motion The Last Pulse option corresponds to a signal processing technique consisting of a time-domain signal of reflected light extracts the "last pulse" corresponding to the farthest object that has caused a detectable reflection of the laser pulse. In this way, the presence of interfering particles, such as snowflakes, along the laser beam does not affect the measurement, and if the laser beam is partially reflected by the edge of an object surface in the foreground, the distance measurement in the background is not affected (within a certain distance) Limit of the difference in the distance of foreground and background areas). These two instruments have a laser measuring beam with an approximate diameter (90% of energy) of about 10-15 mm in the range that is important for this application, and an accuracy of ± 25 mm. The two distance measuring instruments belong to the laser safety class 1 and 1M according to the standard IEC60825-1 (2001). Measurements are available as an analog signal, or through an RS232 serial interface (for model LD90-3100VHS-FLP) or through a parallel ECP standard interface (for model LD90-3100EHS-FLP).

The above-mentioned Optocator ^™ instruments represent a family of triangulation laser sensors that excel in sub-millimeter precision measurement of distances of installation distances of up to 1200mm, with ranges of up to 1024mm, repetition rates of up to 60kHz and a measurement bandwidth of up to at least 20 kHz. Its suitability for this application is limited principally by the potential hazards to the eyesight (unless a fast scan is used, as explained below) and by the small maximum installation distance and range of measurement. 11a and 11b show a possible plan for using a series of fixed laser distance meters (here "FLDM") to perform the required 3DD measurements. For the sake of simplicity, only the LDM units will be used for measurements on one side of the vehicle 11a and an equal number of LDM units should be distributed symmetrically so that the entire carbody is covered. In addition, the number of LDM units in 11a and not necessarily their design as a design, but they have been chosen by the Applicant in order to simplify and support the discussion below. Likewise, for the sake of simplicity, each FLDM is in the drawings, as is unity 481 drawn by a box from which the laser beam 480 is projected through the appropriate LDM optics. The LDM units 481 to 492 , here called "lateral LDM", serve that part of the measuring range which corresponds approximately to one side of the vehicle. The measuring beams, as well 480 , of all lateral LDM have an angle 495 in the vertical direction 496 is inclined. This angle should be chosen to reduce the total number of lateral LDM units, within the desired performance specifications such as sensitivity and resolution, such as minimum sensing distance, and near rail presence as above 10a and 10b was discussed.

A big angle value 495 (where the laser beam almost reaches horizontal orientation) implies the use of a large number of lateral sensors and may cause problems associated with the minimum installation distance between the LDM and the next reference target. A very small angle value 495 is unsuitable, however, because the upper elements of the vehicle body ver the lower parts could darken. As stated above, the greatest 3DD accuracy is desired on the vehicle side in the direction perpendicular to the vehicle side plane, and thus the distance measuring accuracy in this respect is dominant when the angle 495 almost 90 degrees. If, on the other hand, the angle 495 is close to zero, the vertical spacing of the lateral LDM units plays an important role in defining the lateral accuracy in the detection of protruding elements, if this distance leaves significant vertical gaps in the measurement of the surface parallel to the side profile and the rails. It should be noted, however, that at a small angle 495 each projecting element, which must be fixed to one side of the vehicle, should be recognized, at least thanks to its connecting trunk, by a dense series of measurements despite vertical gaps between the laser beams, the result being a limited error in the determination of the length over which the element passes protrudes a vehicle side. Depending on the number of LDM units, according to angle size 495 and according to the physical size of the LDM units (including a weather cover, if required), it may be impossible to install the lateral LDM units in a vertical series with completely overlapping laser beams, and thus it might be convenient to use them to install a shift towards the rails, as is the case for the corresponding groups 497 and 498 the LDM units in 11b is shown. If an LDM unit is used as the lateral FLDM, which has separate front lenses for the laser beam and for the light collection (like the above-mentioned instrument models LD90-3100VHS-FLP and LD90-3100EHS-FLP), then it is recommended to use an optical symmetry plane for to have both planes, which is vertical, and a laser lens that is higher than the detector optics. The use of FLDM units, that is, without any time-dependent control of the laser beam, makes it possible to quite easily meet the condition of continuous measurement in the longitudinal direction. For example, if the LDM laser beam has an effective spot size diameter of about 10 mm at the appropriate pitch, and if the beam is oriented orthogonal to the rail direction on a flat surface parallel to the vehicle side, then a 5 mm longitudinal measurement gap, ie with a safe overlap of adjacent measurements, can be achieved at a vehicle speed of 120 km / h with a repetition rate of about 6600 s ^-1 . The measurement rate for achieving a continuous longitudinal coverage can be reduced by taking the angle value 501 between the laser beam of each lateral LDM unit and the direction 502 of the rails, since this angle is 90 degrees in the assumed situation of the previous set. A low repetition rate allows for the reduction of instrument costs and the choice of one LDM unit from a larger group of commercially available models from different manufacturers.

Two more series of LDM units, as well 493 and 494 in 11a , can be installed to provide adequate coverage of the 3DD measurement range for the top portions of the kinematic gauge line, as long as it conforms to standard UIC profile [050] or GA and GB profiles [053], while another scheme can be easily defined for the boundary line profile GC [053]. Only one LDM unit 499 is in the for simplicity 11b pictured to the FLDM units of the series 493 and 494 to represent. Similar considerations as above for the positioning and orientation of lateral FLDM apply to the optimization of the number of these other series of FLDM units, including the choice of the angle 504 between the laser beam 503 and the longitudinal direction 505 , The direction of the laser beams 500 and 503 in 11b are equal to the vehicle movement, but they could be changed independently by taking the value of angle 501 or from angle 504 replaced by their supplementary angle. The use of three series of LDM units for each vehicle side is not necessarily the best solution and there is considerable room for a free design of the system implementation which also allows further installation criteria for the mechanical structures to which the LDM units are attached. considering the wiring of the LDM units and the electrical safety implications of the presence of the electric train line.

If fixed "time-of-flight" LDM units, such as discussed above used for 3DD measurement, the LDM units must be external so triggered be that the laser beam times for each LDM by a small Time interval can be offset, so that the different units do not interfere with each other.

5.7.6 High Speed Laser Distance Measurement and scanning devices

An alternative to installing a substantial number of fixed LDMs, as discussed above, is the use of a smaller number of high speed laser distance measuring and sampling devices, called HLDSs herein. Many examples and applications for the use of scanning mirrors to control the laser beam and the reflected light in an LDM of the time of passing are well known for performing a series of measurements in different directions with a single laser distance meter. The most commonly used types of control mirrors are prismatic polygon mirrors, a frustum of pyramidal polygon mirrors and tilted mirrors that rotate over a circular base. It is also possible to use tilt mirrors to control the beams, but this solution is not competitive at relatively high sampling rates and in such a case requires scanning in a sine curve in two directions, which is less suitable in this application than a constant angular rate in sampling only one direction. HLDS are often developed for specific applications, but there are some complete models of these instruments on the market, eg the model LMS-Q140i-60/80 [960] from Riegl Laser Measurement Systems in Horn, Austria (for this application, however, would be an individualized version with a higher sampling rate required).

12a and 12b show a possible scheme for using a set of "time-of-flight" HLDS units to perform 3DD measurements on the system 12a shown only half of the HLDS units intended for measurements of some parts on one side of the vehicle and some other parts on the other side of the vehicle. A corresponding HLDS unit should therefore be in a mirror position for the units 520 . 525 . 526 and 527 above a vertical plane of symmetry in the middle of the rail. 12b corresponds to 12a but it's just the only HLDS unit 528 shown the unit 502 corresponds, but has a different orientation.

It is of course desirable to reduce the overall cost and installation complexity for the set of HLDS units, and above all to limit the number of these units, but there are limits here, notably the maximum repetition rate of the LDM used in the HLDS units. Units themselves are used. The scanning angle " 521 which corresponds to the deflection curve of the LDS laser beam is basically determined by the type and characteristics of the control mechanism, such as the number of fronts of a polygon mirror and the orientation of the LDM with respect to the control element. The deflection frequency is determined in the case of the polygonal mirror by the rotational frequency of the axis of the polygon mirror multiplied by the number of mirror fronts. In general, only a fraction of the time can be used for the measurement, corresponding to the angles of rotation of the polygon mirror for which the laser beam has no significant overlap with the edge of two adjacent mirror fronts. Increasing the percentage of the measurement time necessitates larger polygonal mirrors and their size increases even further with the beam size, which is a practical limitation. However, it is possible as an alternative to reduce the spot size of the laser on the mirror fronts with appropriate optics.

When using one of the commercially available LDMs for the time of passing at one of the fastest repetition rates, such as the model LD90-3100EHS-LFP [959], one can as a rule assume that only about 50% of the possible measurements can be carried out, and thus a repetition rate of 12 kHz will be available at each sampling, followed by a dead time with an average sampling rate of 6 kHz. For example, assuming a vehicle speed of 120 km / h and a 40 mm longitudinal measurement gap, this implies a sampling rate of about 833 s ^-1 , and only 7 measurements in this case would be made on each scan. One way to increase the measurement rate is to mount two LDMs in each HLDS unit on the opposite sides of the rotating seal axis, doubling the sampling rate. Their triggering phases can be adjusted so that the series of measurement angles is staggered by the two LDMs, thus compensating for the actual gap in the angle scan of the measurements. More than two LDMs can be mounted in a single HLDS, but overall, the advantage over the use of a larger number of simpler devices becomes smaller and smaller due to the increasing complexity and size of the measurement units. 12a shows that the measuring beams 523 . 523 and 524 at different angles on the vehicle side, and in the comment of 11a and 11b contained comments on the optimization of the angle 495 suggest that a great value for the scanning angle 521 leads to a lower optimization of part of the measuring angle. In addition, the opportunity to perform maximum rate measurements for the LDM used implies that equal angular distances apply to two consecutive measurements, thus leading to a different vertical measurement gap on the vehicle side (in the example of unity 520 ) and also results in a larger vertical opening of the 3DD points on the underside of the vehicle. A further loss of optimization generally comes from including two 3DD measurement ranges with a single LDM unit, each corresponding to a different slope of the corresponding portion of the gauge line profile. The above discussions, especially the fact that its LDM repetition rate means a low sampling rate per sample, are the reason for using fewer different HDLS units, as in 12a each corresponding to a limited scan angle and a certain portion of the gauge line profile. It should be noted that one of the in 12a The criteria used is that the HDLS page that the Measuring window corresponds, always has a negative slope, to reduce the problems associated with the direct impact of rain and the deposition of snow and dust. The positioning and orientation of the four HLDS units 520 . 525 . 526 and 527 in 12a is just one example and the actual number of HLDS units can also be varied.

The comments above 12a and 12b partly refer specifically to HLDS based on the LDM at the time of passing with polygonal mirror scanning, but many of the remarks also apply to other combinations of different LDM units with other scanning systems. For example, the laser triangulation sensors resembling the above-mentioned Optocator ^™ instruments could be used with appropriate mechanical scanning devices provided that the measurement range is large enough and that the laser beam parameters also make the system compatible with train travelers and crew safety criteria for the eyesight , The use of scanning devices for distance measurement is also possible if the beam control is based not on the movement of an optical element, but on other control devices, such as the instrument described in point [043], where acoustic-optical modulators are used.

5.7.7 Super high-speed laser distance measurement and scanning devices

The above considerations on the use of HLDS units based on a maximum measurement rate a few tenths of kHz (thousands of measurements per second) for every Unity, possibly when using more than one LDS for each HLDS. The use of a super-high-speed laser distance measurement and scanning device, here called VLDS, is addressed separately here, as it makes it possible with only a single device complete and adequate sampling of at least half of the entire 3DD measuring range of a vehicle body. Two quite new patent documents [029, 030] represent a very potent one High performance distance measuring device and a fast mechanical scanning system that forms the basis of a Family of VLDS instruments formed by the Zoller & Froehlich GmbH getting produced. The distance is from such instruments measured the modulation phase of an RF-modulated laser beam, aimed at the target and thrown back to the detector of the instrument becomes. The laser distance scanner, without scanning mechanism, has the name "LARA" and is considered a first version with a maximum unique range of about 25 m and a measuring rate of up to 625 kHz available; Furthermore there is a second version with a maximum unique range of about 54 m and a measuring rate of up to 500 kHz. A system for one continuous 360 degrees "vertical Sampling "[030] directs the measuring beam and the beam through a rotating mirror Return beam, which are parallel or coaxial. The most suitable current Models of distance scanners from Zoller & Froehlich for the Use in the system are in the product lines "Profiler" and "IVAR" [961] to find. Such laser scanning devices comprise either the first or the second type of LARA distance measuring devices, wherein the first type for the application is more appropriate because it has a sufficient area, a low laser energy with a safety distance of 1 m from the Eyesight and a 1σ range resolution, which is about 0.8 mm at a measuring rate of 500 kHz, or 0.4 mm at 125 kHz. The linearity error exceeds not 3 mm and a maximum displacement of 1 mm is due to the change the operating temperature of the instrument in the interval from 0 to 40 Centigrade. The measurable minimum range is 0.4 meters and the beam has an average diameter of just under 4 mm in the area the for this application is interesting. The value of the relative reflection intensity also becomes as an output generated by the instrument, and can be used in some applications useful be inclusive the present (for the construction of synthetic 3D images for remote sites are determined and / or appreciated an emissivity value for certain heat release measurements). Due to the detection technology used in this instrument every partial obstacle (for example, dust, snowflakes, the edge of a area in the foreground, which partly touches the laser beam, etc.) a premature radiation reflection to the receiving optics and this leads to reduce the measured distance of the target area in the background. For the same reason, the edge of a surface appears in the foreground farther away than it really is, if any part of the laser radiation reaches an element in the background. This property must be considered the corresponding data processing software become. Furthermore it is recommended to test the instrument regularly by using the distances to some elements of the infrastructure checks to the presence of Dust clouds or intense snowfall to detect the functionality of the system. It is recommended that the measurements be considered invalid if the disorders exceed a certain threshold. Finally, can the appropriate data processing software the presence of a acceptable yet significant disorder (for example, a mild snowfall), when the distance data is filtered more strictly.

13a and 13b indicate a possible practical positioning of a VLDS, which is assumed in the following discussion as the scanning device of Zoller & Froehlich, in a version with a maximum currently detected sampling rate of 18000 rpm [961]. Only one VLDS instrument, namely 540 and 547 , will be in both 13a and 13b shown to perform the 3DD measurements on one side and on most of the upper vehicle surface. A symmetrically positioned partner instrument should therefore be installed on the other side of the vehicle. 13b and 13a are not two orthogonal views of the same installation configuration, because the orientation of the VLDS is different. Assuming that the VLDS instrument is installed in a protective housing, then element is 541 from 13a as well as the element 548 from 13b , the extension of the protective enclosure to reduce the interference of sunlight and the ingress of dust, raindrops and snow to the optical window of the instrument.

13a shows that the installation of the 540 at a relatively high position and above a space of the near rail track, it makes possible more than half of the 3DD measurement range with a total angle 542 between the measuring boundary beams 543 and 546 of just under 60 degrees, and also to achieve an acceptable distribution of the inclination values of the angles between the laser beam and all relevant sections that make up the reference boundary line profile. Using the above sampling rate of 18000 rpm, ie 300 s ^-1 , and the maximum sampling rate, the laser beams diverge from two simultaneous measurements as well 544 and 545 at an angle of about 0.18 degrees, while such a scan resolution angle has a value of 0.43 degrees at a measurement rate of 250 kHz and a value of 0.86 degrees at a measurement rate of 125 kHz.

The maximum sampling rate of 300 s ^-1 corresponds to a displacement of the vehicle between two successive scans of approximately 74 mm at a speed of 80 km / h or 110 mm at a speed of 120 km / h. It is therefore desirable, as discussed above in discussing the 3DD measuring systems, that the angle be 550 between the laser beams 549 and a parallel to the rails is less than 90 degrees, so that narrow mechanical elements that dangerously protrude orthogonally beyond the vehicle side can be better detected. Such a reduction of the angle 550 implies an increase in the angle 542 , which is not a problem, thanks to the excellent maximum angular scanning distance of the specified VLDS. In addition, with reference to 13a , the VLDS should be installed slightly closer to the vertical of the scanned vehicle side, if the angle 550 is reduced to avoid possible darkening of the measurements by passing on the adjacent railroad train. For the in 13a given geometry correspond to the highest values of the vertical clearance in 3DD measurements at the lowest heights of the vehicle body side. Simple geometric calculations show that such a gap, at an angle 550 of about 90 degrees, a maximum measurement rate, a laser beam tilt of about 8 degrees above the vertical, and a position of the 3DD point about 1 meter above the rolling surface is about 110 mm. If the train speed is below the maximum reference in the majority of cases, then an adaptive value of the angle could be 550 Rotating the VLDS about the vertical depending on the train speed may prove to be an appropriate design choice to maximize worst-case compliance between the vertical and longitudinal measurement gaps on the tension side.

5.8 The vehicle body position and calculate its orientation over time

The Performance of the diagnostic methods of detection described below of dimensional and heat-related Defects or dangers of a vehicle body depend on the accuracy of the vehicle Determining the position and orientation of the vehicle body in the period in which the measurements are carried out for the car body itself. The following is a method for determining the "VBPO", for "Vehicle Body Position and Orientation ", which, as explained in detail below is, the position and orientation in the space of a rail vehicle body as express a function of time, when the vehicle model has been detected and the required information are available in the vehicle database.

14 shows a generic rail vehicle 250 with two articulated bogies, each with two wheelsets, on the rails 251 and 252 moves. A three-dimensional Cartesian coordinate system C _VB integral with the car body 250 and centered on 257 with the coordinate axes X _VB 260 , Y _VB 255 and Z _VB 258 is considered here and allows to determine vectors or the position of any point on the car body, regardless of the position and orientation of the vehicle itself. Although in principle the position and orientation of the coordinate system C _VB could be chosen arbitrarily for the car body, recommends That is, and it will be assumed in the following discussion that the plane X _VB Z _VB coincides with the plane of symmetry of the vehicle body, and that the zero point of the axis Z _VB coincides with the zero assumed definitions in the LDF function. The axis Z _VB should pa be parallel to the rolling surface and refer to a new vehicle resting in the middle of the rail. It is also assumed that C _VB matches the coordinate system used to define all relevant vectors based on the car body in the vehicle database. A particular choice for the height of the zero point of C _VB over the rolling surface (of a new vehicle, ie without any wear) may be coincidence with the roll center, as defined in UIC Leaflet 505-1 [050]. However, many other options for positioning C _VB with respect to the body are also possible.

In addition, a three-dimensional Cartesian C _GB coordinate system becomes integral with the ground and centered 253 with the coordinate axes X _GB 256 , Y _GB 259 and Z _GB 254 taken into consideration. The coordinate system C _GB allows the assignment of vectors for any element on the ground, and above all the assignment of positions and orientations to sensors and instruments. element 264 represents the reference center of an optical measuring instrument, and the axes 265 . 266 and 267 belong to a coordinate system C _MS for this particular instrument. The laser distance measuring devices described above directly measure the position of the felt element as 3DD 263 in their own coordinate system C _MS and its measurement uncertainties are normally defined first in this same coordinate system. Other sensors, such as optical barriers, VIS and NIR imaging devices, and IR imaging sensors, assign graded measured values to a versor and a vector origin in their own coordinate systems C _MS , since they can not resolve the distance of the sensed element. In the case of these last sensors, their C _MS coordinate systems are those normally used to define the uncertainty parameters, such as MTF or a section of an optical beam for the distance. Coordinate transformation formulas, such as those discussed below, involving coordinate systems C _GB and C _VB , and which relate to calibration parameters, allow measurement vectors and measurements to be expressed for each instrument in relation to values, such as vector uncertainties, in coordinate system C _GB . The orientation of the three axes C _GB towards the ground and especially with respect to the local orientation of the rails could in principle be arbitrary, but it could be more practical, and it is also assumed here that the axis Z _GB SMI runs parallel to the rails and that the Axle X _{GB is} perpendicular to the rolling surface, assuming that the corresponding railway line is straight. The axis X _GB is therefore not vertical and the inclination of the rail is not equal to zero. The axis _GB is therefore parallel to the rolling surface and perpendicular to the rails. According to the use of the ZDF function, as explained below, the position of the zero point of C _GB along the rails may coincide with the 0 of the axis L as defined above. The distance of the origin of C _GB from the plane vertical to the rail axis results from the calibration process, which will be discussed below. The height of the zero point C _GB over the rolling surface is also arbitrary and one can assume that it is set to zero during the calibration of the system. In practice, this height will change over time due to wear of the railhead and rail reduction, and this will be taken into account in direct or indirect measurements, as explained in section 0 below.

The VBPO can be defined as a vector function of time, which in the space C _{GB is} the position of the center C _VB 257 and the rotation angle of the coordinate system expresses. The VBPO uniquely corresponds to a coordinate transformation function Ω (t), depending on the time t, thanks to which in vehicle-based coordinates C _VB the measurement vectors defined by the ground-based coordinates C _{GB are given} by the following formula V VB = Ω (t) V GB (113) where V _{GB is} a vector in the ground-based coordinate system C _GB , and V _{VB is} the same vector in a vehicle-based coordinate system C _VB . When using the RPY (Roll-Pitch-Yaw) transformation convention, this function can be expressed by the following formula

wo θ(t), φ(t) und ψ(t) die Roll, Pitch- und Yaw-Winkel, und X(t), Y(t) und Z(t) die Übertragungskomponenten von C_VB zu C_GB sind, die zusammen einen Ausdruck der VBPO-Funktion darstellen.For a combined rotation transfer transformation in homogeneous coordinates, where the time-dependent homogeneous transformation template corresponding to the function Ω (t), is from the combination [Ω (t)] = [Ω RZ (T)] [Ω RY (T)] [Ω RX (T)] [Ω LD (t)] (115) the individual matrices for transmission as well as the RPY rotations result

where θ (t), φ (t) and ψ (t) are the roll, pitch and yaw angles, and X (t), Y (t) and Z (t) are the transmission components from C _VB to C _GB , which together represent an expression of the VBPO function.

sin(τ) = τ (121) und

wo τ für jeden beliebigen θ, φ und ψ-Winkel ausgedrückt in Radianten steht, woraus eine bedeutende Einsparung der Berechnungszeit bei den unten erörterten Berechnungen resultiert.Because of the small maximum values of θ, φ and ψ (using the orientations determined for the above coordinate systems C _GB and C _VB ), their sine and cosine functions can be easily replaced in the Ω expression by the truncated ones Formulas of the expansion series

sin (τ) = τ (121) and

where τ is expressed in radians expressed in terms of any θ, φ and, angle, resulting in a significant saving in computational time in the calculations discussed below.

The conversion of vectors from the vehicle-based coordinate system C _VB to the ground-based coordinate system C _GB can be done using the following formula V GB = Ω -1 (T) V VB (123) where Ω ^{-1 represents} the inverse function of function Ω. The commutative property does not hold for the matrix calculators in Equation 115, and the template corresponding to Ω ^-1 is given by the following expression [Ω -1 (t)] = [Ω -1 LD (T)] [Ω -1 RX (T)] [Ω -1 RY (T)] [Ω -1 RZ (T)]. (124)

The solution of the VBPO computational problem directly attributable to solving the conversion of vectors Ground coordinates in vehicle coordinates, and vice versa, leads, passes Thus, in the determination of the functions θ (t), φ (t), ψ (t), X (t), Y (t) and Z (t) of Time t for a certain car body. This requires an expression for each of these Functions selected which includes some parameters that can be optimized by to minimize a function that determines the degree of correspondence between a Series of 3DD measurements on a vehicle and some known ones Characteristics of a corresponding rail vehicle model expresses because such features in the vehicle database, after vehicle recognition, available are.

The VBPO calculation is accomplished through an iterative process that assumes initial "guessed values" for the parameters that may change in the solution discovery process. It is obvious that a limited number of parameters to be optimized as well as an appropriate choice of the first estimates for the parameter values are important in order to find the sought solution faster and more reliably. An important simplification in the VBPO calculation problem arises from the use of the LDF function calculated for the corresponding vehicle in vehicle recognition, as described above. In fact, assuming the above C _VB coordinate system, the function Z (t) may be considered equal to the LDF function L (t). Although in principle Z (t) could be specified by a VBPO estimation process, the applicant prefers to consider them equal to the corresponding LDF, as this makes the VBPO calculation method faster, easier, and safer. The game of the vehicle body in the direction Z _VB with respect to the position of the Radsatzmitten or the position of the axes of rotation of the bogies 261 and 262 is in reality usually very small, and therefore the use of the 3DD data to more closely determine the Z (t) estimate obtained by the LDF is of little help if the LDF is carefully selected with a sufficiently large number of suitable ones Wheel sensors and a suitable calculation method was performed. The choice made for the orientation of the X- _GB axis allows X (t) to be considered time-independent, unless very accurate 3DD measurements are made. The initial estimate of the value X ₀ may be derived from the height of C _VB over the rolling surface available in the database. The estimated value for the gap angle φ ₀ can be set to zero and this angle can be considered constant because of its actual little change in rail vehicles, unless very accurate 3DD measurements are made. The lateral shift functions Y (t), θ (t) for the roll fluctuations, and ψ (t) for the Yaw fluctuations are the main objective of the VBPO determination procedure since they represent the values that mainly determine the lateral displacement of the vehicle body.

in der die Mengen ζ_r das Ausmaß der Positionsübereinstimmung eines bestimmten Fahrzeugs aus der Fahrzeugdatenbank mit einer oder mehrerer 3DD-Messugnen ausdrücken, und in der die Werte σ_r die entsprechenden Standardabweisungen darstellen. Die Abhängigkeit der Werte R ζ_r von den zu optimierenden Parametern zur Definition der VBPO-Komponenten ergibt sich aus der Verwendung der Ω oder Ω^–1-Umwandlung zur Berechnung der Werte ζ_r in den Koordinatensystemen C_VB oder C_GB.The set ζ 2 / VBPO, which is minimized to determine the angular and shift components of the FBPO function, can be expressed by the following chi-square formula

in which the quantities ζ _r the degree of positional conformity of a particular vehicle from the driving express the database of witnesses with one or more 3DD measurements, and in which the values σ _{r represent} the corresponding standard rejections. The dependence of the values R ζ _r on the parameters to be optimized for the definition of the VBPO components results from the use of Ω or Ω ^-1 conversion for calculating the values ζ _r in the coordinate systems C _VB or C _GB .

The Applicant has considered a few other definitions and corresponding calculation methods for the values ζ _r and a brief summary of them follows below, their importance in determining the calculation times, the safety of the method and the effects on the preparation of the corresponding characteristic data, which are described in the vehicle database are stored, sets out. Such options are mainly related to the choice of vehicle features, the measurement system used to determine the 3DD values, and the definition of ζ _r .

In general, if the search for the VBPO solution is started with initial estimates for the VBPO components, the corresponding initial error is θ (t), φ (t), ψ (t), X (t), Y ( t), ie of all VBPO components except Z (t), is a very important aspect to consider when defining and calculating ζ _r .

If the 3DD measurements are performed using one of the means described above using either fixed or scanning distance laser meters, then the gap between the 3DD measurement directions and the possible gaps between measurement beams together can be considered as the main constraint in the calculation of the values ζ _r .

A first type of vehicle element that Applicant recommends to use in this case is a plane surface, here called a "large plane surface" or "F1", using the distance of the 3DD points from the surface as a definition of ζ _r , If more than one 3DD measurement from the LDM sensors relates to the same flat surface, then ζ _r can be defined as the square root of the sum of the individual Qudrat distances. This element can easily be encoded in the vehicle database by a number of parameters, such as the three vertices that define a rectangle in coordinate space C _VB . An important advantage of this choice is that one or more of the corresponding 3DD points can be easily selected using a sufficiently long and wide area that ensures such points relate to the planar area element, with the maximum possible error in the initial values of the VBPO components. Some examples of F1 elements are a flat part on the side wall of a wagon, a part of a flat wagon roof, a flat inclined part of an upper cover of a wagon for transporting coal, or a flat plate for attaching labels to a chemical tanker wagon. If these flat parts are parallel or nearly parallel to the plane X _VB Z _VB , then the distance to a 3DD point to be hit is particularly effective for defining the side shift of the components Y (t) of the VBPO function, while the effect on the definition of roll component θ (t) is minimal when the surface height almost reaches the height of the roll center, and the effect increases as the difference between the two heights increases. The effect on the Yaw component ψ (t) will usually be large unless the element is close to the X _VB Y _VB plane. If, conversely, the flat elements are parallel or nearly parallel to the plane Y _VB Z _VB , then the distance to a 3DD point to be hit is particularly effective for the definition of the gap angle φ (t) and that of the components of the vertical Displacement X (t) (or only X) of the VBPO function, while the effect on the definition of the roll components θ (t) is minimal when the flat element is close to the plane X _VB Z _VB , ie near the vertical of the roll center , and it increases with an increasing lateral shift from the X _VB Z _VB plane. If a flat element is parallel or nearly parallel to the axis Z _VB and forms an angle with the axis X _{VB that} is relatively far from all multiples of π / 2 radians, it will generally affect X (t) and Y (t) while affecting θ (t), φ (t) and ψ (t) depending on its position over the vehicles. Similar considerations may also be made for other orientations of a flat element, taking into account the fact that in this case θ _{r is} related to the position and orientation of the vehicle body over a distance measured in the orthogonal direction to the surface of the flat element. The relative importance of a particular flat element in determining the VBPO components, of course, depends on the uncertainty of σ _r with respect to ζ _r , a value that must be calculated, and usually on the orientation and position on the flat element depends.

A narrow and elongated flat surface is the second practical solution that is compatible with LDM 3DD measurements, and is referred to herein as a "thin flat element" or "F2" element. This element, such as F1, can easily be coded in the vehicle database as a series of parameters, such as the three vertices that define a rectangle in coordinate space C _VB . This element is relatively light even in the important and relatively difficult case of flat rail vehicles, as it may, for example, correspond to the side surface of parts of the loading deck on a wagon. In this case, it is expected that only a fraction of the LDM measurement points selected with the certain initial uncertainty in the LVPO function will be valid, in the sense that they provide sufficient intersection of the cut surface of the laser beam and the thin surface of the element when measuring, depending on the type of LDM. In this case, there should be a preparatory process to filter the candidate 3DD measurement points. Such filtering is particularly easy if the F2 element has a clear "three-dimensional contrast" to the background, ie if there is a significant distance in the direction of the measurement laser beam between the measured distance values at the edges of the element and the background distances LDM instruments, especially those based on a modulation phase, provide highly diffuse measurements when the light reflected back from LDM comes partly from the edge of the element and partly from the background, and therefore the filtering process should be able to match those 3DD points For example, this filtering of the candidate candidate 3DD points for an F2 element may be based on the search for a 3DD point subset that matches a plane with a minimum squared error whose Position and orientation are limited around the F2 element position and orientation, since the Also, elements of the candidate subset must be within a rectangle that can be considered the element itself that has been scaled down to accommodate the size of the cross section of the measurement beam and the orientation of the beam relative to the element surface. One possible way to compute ζ _r for an F2 element is the same as described for F1, using only the points found by the above filtering process. Another way could be to define the plane that interpolates these points, and then define ζ _r by the average squared distance between the element and that plane.

A plate "cut-out" from a round surface, here called a "round plate element" or "F3" element, may be useful in certain types of freight rail cars, such as chemical tankers. This element may be defined in the vehicle database by a few parameters, such as the origin and direction of the cylinder axis plus the cylinder radius and planes defining the plate, or the coordinates defining the two linear plate edges and a line parallel to them. which is equidistant from either edge, or otherwise with few parameters that must be defined such that the calculation of ζ _r . is relieved. The size of the F3 disk can be presumed to be large enough to allow the use of a simple calculation of ζ _r , as described above for F1. The features F3 are attractive as long as they are large enough, since they influence both the VBPO components X (t) and Y (t) and, at the same time, the components θ (t), ψ (t) in the example of the horizontal cylinder in that the longitudinal distance of the element from the plane X _VB Y _{VB is} sufficiently large.

A linear solid structure, called "rod-like element" or "F4" element, is another option that is compatible with the LDM 3DD measurements. This element can be used, for example, for handrails of catwalks on railway tankers and for other external horizontal, vertical or inclined elements which consist of a tube and constitute a fixed component of the vehicle model. This element can be coded in the vehicle database by two points of the rod axis at its ends and by the rod radius. Preparatory processing of the element data as in the case of the element F2 is recommended, and a similar procedure can be used for filtering the 3DD measurement point candidates. ζ _r can be calculated as the square root of the sum of the squared distances between selected points and the rod surface, or by defining the line that interpolates these points, and defining ζ _r by the average squared distance between the corner line of the element and such a line.

A V-shaped structure is another possible element that is compatible with the LDM 3DD measurements, here called "V-shaped element" or "F5" element. This element is widely used in many cases and, like Type F2, can be very useful in relatively difficult cases such as flat rail vehicles. Preliminary processing of the 3DD data is also recommended in this case by filtering the data, which in this case results in the classification of selected points into two groups corresponding to the two V-shaped planes. ζ _r can be calculated here as the square root of the sum of the distances between selected points and the corresponding surface, or by defining the V-shaped corner line by interpolating the two planes to define ζ _r by the average squared distance between the element's axis and lead to such a line. This type of element is characterized by a relatively complex preparatory processing process, but it provides a definition of ζ _r that is very crucial to the definition of min at least two critical VBPO components, depending on the orientation and position of the element itself.

A slot in a flat surface is a particular type of element, here called a "slot element" or "F6" element, which may be compatible with the LDM 3DD measurements. This element can be encoded in the vehicle database as the slit width and the two endpoints of its centerline, along with the encoding of the corresponding surface as described for the element of type F1. In this case, the candidate data points should be filtered to select the 3DD points associated with the slot by measuring the non-surface compatible distances and by finding three-dimensional coordinates and measurement vectors with the linear slit to match. In this case, ζ _r can be defined by the interpolation of the slot centerline and the calculation of the square root of the average squared distance of its centerline to the centerline of the element as encoded in the vehicle database. The discarded data points of the surface far enough away from the slit edges can also be used to define a second value ζ _r calculated in the same way as for the element F1.

A linear and low ridge-like structure over a flat surface can also be compatible with LDM 3DD measurements, provided they are accurate enough. This structure corresponds to an element referred to herein as a "linear ridge element" or "F7" element. The definition, the preparatory processing method and the definition of ζ _r are similar to those explained above for element F6.

A plate cut out of a spherical surface, here called a "spherical plate element" or "F8" element, may be suitable as an element for some special cases of rail vehicles when LDM 3DD measurements are used to define ζ _r . The definitions and methods for preparing the data and calculating ζ _r are conceptually similar to those given above for F3.

trihedral Elements with sufficiently large ones Plains, here "trihedral Called "elements" or "F9" elements, can treated as an obvious extension of the F5 elements and are an information-rich option using the LDM 3DD measurements is compatible.

A principal alternative to LDM 3DD measurements for calculating the VBPO function is the use of stereoscopic image processing for appropriate elements stored in the vehicle database. In fact, even if the applicant has expressed above his concerns about the safety of this method for the detection of protruding obstacles, he considers this method to be an interesting option if the element to be located in the room can be chosen in a suitable way, such as in the case of items stored in the vehicle database. It is of course possible to install additional VIS or NIR imaging sensors with the specific purpose of performing 3DD measurements on selected vehicle elements, but Applicant prefers to use as much as possible the same imaging devices installed as described above for this purpose to determine the vehicle images used as input to the OCR process. In view of the need to simultaneously map features to at least two different imaging devices to associate the vehicle elements with stereoscopic image processing localization, and because the candidate elements are often positioned at outer positions around the vehicle, one could use a larger number of linear imaging devices 10a suggest. The use of further angles and a higher number of image processing elements in the arrangement for these image forming apparatuses is a solution that would make it possible to prevent a significant increase in the number of linear image forming apparatuses by using the central portion of the linear image forming apparatus for OCR and documentation purposes. and the side panels for stereoscopic image processing, taking into account that for the latter process, in some cases, a lower image quality than OCR is acceptable in the method. A large shift between the imaging devices results in the definition of the position of the elements by stereoscopic image processing becoming a valid option for a limited set of elements, including particular lines and shapes lying on nearly flat surfaces. However, large displacement of the image forming apparatus is characterized by greater accuracy in measuring the distance of the element from the image forming apparatuses. However, the installation of a second imaging device only for stereoscopic vision close to an image forming apparatus that is responsible for the two-dimensional image processing is an option that can be considered when the stereoscopic vision as a primary method for the detection of vehicle characteristics in the context of Recognition procedure of the VBPO function components is used. Another approach to the use of imaging devices in determining the position of vehicle elements in space is the processing of individual two-dimensional images (including images that constructed by data from the linear imaging apparatus) containing vehicle elements for which one or more size measurements can be made. The use of this latter method generally does not imply a change in the number of linear imaging devices or their characteristics with respect to the discussion of the imaging devices in FIG 11a and 11b for OCR and documentation purposes.

An element that corresponds to the presence of an undefined visual character or mark on a known flat surface is called a "flat undefined visual element" or "F10" element Rail vehicles may be used on which marks or drawings are expected on a flat surface, such as a part of a rail vehicle wall.The surface coding given above for the F1 elements can also be used in this case.If this element is used in the VBPO method In most cases, a processing procedure for stereoscopic images produces a localization by creating a set of 3DD points that defines the area, and the corresponding value ζ _r can be expressed as the square root of the sum of the squared distances of the corresponding 3DD points the surface of the element, as in Element F1 Alternatively, the relevant plane in space C _{GB may be defined} via the image processing results, and the square root of the average squared distance from that plane from the corresponding surface defined in the vehicle database may serve as a definition for ζ _r .

A visual form with known dimensions and an unknown exact position on a flat surface is called here "fixed-size floating" or "F11". The processing of the corresponding image or images in this case can be done by a method of processing stereo or single images, thanks to the absolute sizes of the shape. The image processing results locate the corresponding plane in space C _GB and the square root of the average squared distance from that plane from the corresponding surface defined in the vehicle database can serve as a definition for ζ _r .

If a visual form on a vehicle surface has a defined size and a defined orientation and position, it corresponds to a "completely defined shape" or "F12". The processing of the relevant image or images may be done in this case by a method for processing stereo or single images, and the image processing results may in this case for calculating _r ζ of the square root of the average squared distances between two forms are calculated by the measurements and the corresponding codes are defined in the vehicle database.

Not all of the above types of elements must necessarily be implemented in a system form. Other types of elements may be defined together with their encoding in the vehicle database and by one or more calculation methods of one or more values ζ _r . The methods for quantifying the uncertainties σ _r for the values ζ _r are not discussed here, as they would require very many lines of text, and because experts, such as experts in numerical computation in robotics, can define them without much difficulty.

in the Generally, more than one element for a particular vehicle model used in the vehicle database and the choice of elements takes place based on a number of different criteria with the aim of to maximize the performance of the VBPO procedure while keeping the complexity and the Cost of the population the vehicle database limits. The applicability of a particular element obviously depends on the geometric characteristics and of the operational characteristics of individual Vehicles depend on and are determined by the type of sensors used in the system implementation be installed, as well as determined by their position and orientation. If more than one version of the system has been implemented (with different instrument types and / or their positioning), the vehicle database can exist as a single version and Contain elements that are applied only by some of these versions can be. A Set of such basic criteria is defined here below to decide which elements are more appropriate to the contents of the vehicle database for a defining a particular vehicle without implying that its Order corresponds to their relative importance, or that others Criteria that are at least as important do not exist.

An obvious but fundamental criterion in choosing the positions and the size of the vehicle elements for a vehicle model is the prevention of disturbances from the load, and above all the presence of imaginable abnormal loads and possibly too far protruding objects caused by the system would be recognized. However, this criterion should not apply to all elements It should be disregarded and incompatible with the possibility of defining sufficiently well-known elements on the basis of the other criteria discussed below.

The different types of elements have different effects the approaching the functions θ (t), φ (t), ψ (t), X (t) and Y (t) off, too dependent from the positions of the elements. The types and positions of the selected elements should therefore possible no undecayed VBPOP components, especially not the critical ones.

The functions θ (t), φ (t), ψ (t), X (t), and Y (t) are approximated by measurements made at a number of times at corresponding different vehicle offsets with respect to the instruments installed on the ground become. The choice of elements and their position on the vehicle should therefore take into account the spacing between the elements and the positioning of the elements to achieve a sufficiently uniform distribution of the registered elements in the vehicle scan for effective determination of the time-dependent VBPO components, particularly the rather critical components. The twelve types of elements described above are clearly characterized by a different complexity and a different degree of use with computing means. The required maturity of calculation tools should therefore be taken into account, taking into account the fact that some of the calculations, especially the processes referred to as preparatory calculations when annotating the various elements, are executed only once, whereas the values ζ _r be calculated several times in the minimization of expression 125. The calculation of the VBPO function should also be taken into account when deciding the longitudinal position of the sensors around the rail, in particular as regards the performance of the gauge line profile and the thermal diagnostic methods for the body. A first consideration here is that the calculation of the VBPO, and especially of the components φ (t), ψ (t), X (t), and Y (t) by the measurements given above, requires that appropriate 3DD data be reasonably short Time intervals are available for at least two places with a sufficiently large longitudinal spacing. A second consideration is that, similar to what said above about the positioning of the wheel sensors in relation to the SMI, it is important in this case that the three-dimensional position and orientation of the vehicle body are known with sufficient accuracy when measuring in the vehicle With respect to the gauge line or thermal detection, consideration must also be given to the time difference or longitudinal displacement between the measurements for the VBPO calculation and the measurements associated with the vehicle components in the transformation of Ω or Ω ^-1 .

As discussed below will require the use of 3DD measurements for the functions in the context of the gauge line profile, that these at least executed twice be a false alarm due to flying disturbing elements rejected and it was said above that the sensors positioned at the SMI for the Measurement of 3DD data of the two lateral sides of the vehicle body advantageous in the longitudinal direction should be placed. Therefore, an advantageous place could be for the relevant ones 3DD measuring sensors will be an almost equal distance along the SMI, where you alternate their installation on the two rail sides and a big enough one Distance between their two extreme positions in relation to the Positions of the sensors for the functions of boundary line profile and thermal diagnostics maintains, because these sensors partially coincide with the sensors performing the VBPO calculation.

As in the case of the LDF functions discussed above, the VBPO functions, and especially the functions of θ (t), ψ (t) and Y (t) and Z (t) can be approximated by using cubic spline functions. In such a case, the number of spline pieces and their limitations depend on the train speed and to the actual Dynamics of rail vehicles fit, assuming the worst Case or, if possible, should apply some parameters of the vehicle file. Using the truncated Fourier series or their combination with the spline functions could represent an alternative, especially in the case of VBPO rotational components.

Multi-parameter standard techniques can be used to optimize the VBPO definition while minimizing expression. The choice of calculating ζ _r in the coordinate spaces C _GB or C _GB remains open, and one can apply either the Ω transformation of 3DD measurements or the Ω ^-1 transformation of coordinates of the vehicle elements. The second choice may be advantageous if the element is described by a few vector parameters, while there are many vectors for the relevant 3DD measurement parameters. Because of the possibility of interfering with the measurement of an element, especially with a normal or abnormal load element, the VBPO calculation procedure should be such that, at least for convergence problems, the calculation is repeated, and one or more of the considered elements are neglected until satisfactory convergence is achieved. Failure to achieve satisfactory convergence of the VBPO calculation procedure would result in the generation of an error flag or an error indication which must be adequately addressed by the other functions available to the corresponding vehicle.

5.9 detection of dangerous Deviations from the boundary line profile of the body and the load of an identified vehicle

The method that 237 in 3 is described below and is used to detect hazards in the gauge line profile of rolling stock whose model has been identified and whose BVPO function components have been calculated as explained above.

The corresponding hazards in the gauge line include vehicles, that are not compatible with the gauge line profile, invalid load profiles due to inappropriate cargo or a freight shift or an irregular cargo. This method does not apply to the lower vehicle parts, as defined by the UIC series 505 leaflets [050, 051, 052], since such parts are subject of another discussion later in this Document are.

15a resembles the 5 in UIC Leaflet 505-1 [050] and includes some profiles and parameters given in the discussion below on hazard recognition in the vehicle body gauge and its cargo. The applicant clarifies that the following comments apply to 15a as well as the comments too 15b below, should not constitute an authentic and complete interpretation of the relevant contents of the UIC Code, and that they serve only to improve the reading comprehension of this section of the document, assuming that the reader sufficiently satisfies the UIC-505 series of Leaflets knows. The two axes 560 and 561 define the "normal coordinate system" for a rail and for a transverse vehicle section, which is perpendicular to the longitudinal axis of the rail center axis. These normal coordinates, used in the leaflet series UIC 505 , apply to both the vehicle and the track, assuming that the vehicle is stationary and symmetrically positioned such that its vertical axis passes through the central axis of the track in situ. The zero point of both axes 560 and 561 lies on the rolling surface at the same distance from the rails. The overall theme of the gauge line profile of a rail vehicle, in accordance with the UIC Code, is based on the assumption of a particular gauge line profile that includes a reference profile and a set of rules that, by taking such a profile as a common foundation, allow rail vehicle services, define a maximum allowable profile for the vehicles (and their freight) as well as allow the route and work services to define a boundary profile for the infrastructure elements. The different profiles 562 . 563 . 564 . 571 . 572 and 574 are through the series 505 the UIC leaflet defines the classification of the effects of a number of factors, such as the position of the vehicle body parts relative to the rails in a curve, the inclination of the vehicle about its center of roll, the lateral displacement of the entire vehicle body due to a strong or too strong tipping, the kinematics of bogies and axles, the defects of the railway line, the sagging of rails and sleepers in track ballast, etc. The profile 562 corresponds to the limiting gauge line profile in the vehicle design and defines the maximum displacement positions from the axle 560 from any vehicle part (at a certain longitudinal position on the vehicle). The profile 563 corresponds to the reference outline of the kinematic gauge line profile of the vehicle and the distance E 566 between this outline and the vehicle design profile 562 is given by the "reductions" which must be assessed in accordance with the rules of UIC-505-1 Leaflet [050]. In fact, the quantity E corresponds to either E _i or E _a , depending on the position of the relevant transverse portion lying between the first and last axles not mounted on bogies, or between bogie boxes or outside this position portion. The component z of E is the "quasi-static lateral displacement" corresponding to the lateral inclination resulting from the component of the asymmetry angle of the vehicle that is one degree too large and the partial effect of the too-strong or too weak-tilting. The profile 564 corresponds to the outer limit of any vehicle part as viewed from the reduction formulas. The profile 564 is from the profile 563 by the distance S or the "lateral projecting" 567 separated and different from the profile 562 by D, the distance of the lateral displacement in total 568 equivalent. S is the amount by which the vehicle outline exceeds the reference profile when cornering and / or exceeding the rail gauge line by 1435 mm (standard rail gauge line profile). The half width of the vehicle profile at a certain height plus the amount D and minus half the width of the datum line profile at the same height is equal to the effective value S with respect to the datum line profile.

The reductions E _i or E _a must be equal to or greater than the amount DS ₀ , where S _{0 is} equal to the maximum value of S, to preclude any vehicle part from being outside the vehicle position limit profile. The profile 571 corresponds to the kinematic obstacle profile of the vehicle and its half width exceeds half the width of the profile 564 by a distance 569 which belongs to the quasi-static displacement that is not taken into account within D. The profile 572 corresponds to the limit positions of any part of the route, and is determined by the profile 571 through the distance 570 separated, which corresponds to the fluctuations and the asymmetry below a degree and reflects the lateral shifts due to rail defects. The profile 573 corresponds to the actual physical boundary outline of the infrastructure and its half width exceeds the profile 572 around the distance 574 chosen for a particular rail, taking into account special operations or situations such as the transport of wide and / or very long loads or the frequent occurrence of strong crosswinds. The limit profile of the vehicle position 564 separates the competences of rail vehicle services from those of the route and work services that are responsible for the clearance in the dashed area of 15a are responsible.

A major aspect important to this invention is that the approved vehicle profile 562 that is, its maximum permitted width at a certain height above the rolling surface, depends on the distance between the corresponding transverse section and the bogie housings or the two outer fixed axes. This results from the fact that the reductions of the vehicle relative to the reference boundary line profile include the "geometric lateral shift" of the vehicle body at a radius of the rail turn equal to or greater than 250 meters, with the additional side clearance for a smaller turn radius in determining the required Gap between the infrastructure profile and the limit profile of the vehicle position. 15b shows, similar to 4 UIC leaflet 505-1 [050], a horizontal section of the vehicle with the different parts 577 . 583 and 582 the profile of the maximum design width of a vehicle, and such a profile is significantly influenced by the geometric lateral displacement of the position of the transverse section along the longitudinal axis 575 depends. In practice, most vehicles have a design profile 585 that is the dashed area 580 Corresponds to a rectangle that is within the maximum allowable profile, largely due to the constraints imposed by the geometry of station platforms and other structures, and that at the four corners 576 is bevelled.

The UIC code leaflets in terms of Of course, the vehicles and the infrastructure profiles also determine Profile boundaries for the vehicle loads and they are also exceptional in case of Freights used on wagons. The freight profiles are however subject other standards, such as, if applicable, the vehicle freight regulations in the RIV Agreement [060] and its annexes. Especially Part 5, Volume I, Annex II of the RIV Agreement gives a number of limits to the cross-sectional outline of freights based on a set of "cargo boundary profiles", the reductions according to a Group of applicable tables. In the same section will be other cargo geometry boundaries in terms of the extreme longitudinal Charging position specified and it contains information on the use of compound, multiple and articulated cars.

part 5 of the RIV Agreement [060] (and also UIC Leaflet 502 [075]) deals with coding and the labeling of special loads, the corresponding coding of train lines and the rules to be followed in this coding to ensure the safe transport of such special loads. The encoded special freight profiles also represent a type of profile which can be used by the system. UIC Leaflets 596-5, 596-6 and 597 [054, 055, 056] discuss the profiles for combined Rail transportation. The system also uses these profiles in its Functions for detecting dangerous Deviations from the boundary line profile. In the case of a combined transport and in the case of a transport of special loads ("special deliveries" in the UIC leaflet 502 [075] and elsewhere), the system can use the OCR and OCR-like Processing of pictures of a vehicle, the special markings, which define these norms, read. Reading these marks allows the system to determine which particular profiles are used to prevent injuries to the profiles of combined transport or the Recognize profiles of special freights.

The detection of a violation of a loading profile but not of a maximum permissible profile of the vehicle and its cargo may be an indication of an element that has slipped out of its correct position and will continue to slip at a later time, thus becoming a danger of the gauge line profile can. In addition to the UIC design profiles and load profiles, it may be useful to consider the actual design profiles of wagons in the practice of this invention, for example, because the system provides, if desired, the presence of abnormal elements could recognize within an approved vehicle profile, such as a black passenger lying in free space under the tank of a chemical vehicle transport rail vehicle.

16 is used below to identify the method of detecting the presence of an element beyond a particular profile, integral with the passing vehicle. Such a relevant profile may be the permissible design and loading profile in accordance with the principles of applicable UIC 505-5 [052] or it may be the design profile known from the vehicle database (eg in accordance with the applicable Annexes of the RIV Agreement [060] and / or UIC Leaflets 596-5, 596-6 and 597 [054, 055, 056]). The axes 601 . 602 and 603 correspond to the ground-based coordinate space C _GB with the zero point 600 while the axes 605 . 606 and 607 to the vehicle-based coordinate space C _VB with the zero point 604 belong. C _GB and C _VB are the same coordinate systems used in the commentary above 14 despite the differences in their respective positions and orientations, the relative rotation of the radian π about the axis X of one of the two coordinate systems resulting from a different traction direction of 16 and 14 result. The reference numbers 608 to 612 stand for four profiles defined over different transverse sections of the vehicle, corresponding to different values of the longitudinal coordinate of the vehicle z _VB indicated by the thick point marked intersections of the vertical profile axis with the axis Z _VB . These profiles can be used in general for vehicle sections such as 578 and 581 in 15b or belong to other profiles mentioned above. The box 614 represents an optical device for distance measurement, which is a measurement for the position of a point M, indicated by 617 , indicating on the surface that can be assumed as an element of the vehicle. 616 shows the measured vector with the origin at the position 615 at the meter, which by the coordinate values 618 . 619 and 620 is defined relative to the axes C _GB . The coordinate transformation arithmetic Ω, whose time-related parameters for the vehicle were calculated as explained above, is expressed in Equation 113 to define the position of M through the values 621 . 622 and 623 used in the coordinate system C _VB . Ω can also be used to adjust the measurement vector to the coordinate system C _VB , which can be used to estimate the corresponding measurement uncertainties for the position in different directions. The comparison can therefore be made between the position 616 of the point M and the envelope defined by a set of bounding profiles in the same coordinate system C _VB . However, it will be apparent to those skilled in the art that the computations related to the detection of elements beyond certain vehicle profiles may alternatively be performed in the coordinate system C _GB using the Ω ^-1 transformation applied to the corresponding vehicle profile is defined in the coordinate space C _VB .

Any type of vehicle-based pre-defined profile for detecting elements at profile violating positions may be stored in the vehicle database using a suitable longitudinal space, such as 613 , and the margin distance for comparing the detected 616 can then from the distances 624 and 625 , or from multiple distances to accommodate a non-linearity, at the appropriate height of the next profile 610 and 610 be interpolated. The gap distance 613 can be defined as a fixed value, or it can depend on the longitudinal position and ideally should be chosen taking into account the first and second derivative of the profile width with respect to z _VB . An obvious alternative way of defining the profile is to use a series of profile lines corresponding to horizontal (rather than vertical) surfaces. Alternatively, the vehicle profiles may be stored as a set of surfaces or calculated with parameters and appropriate formulas, with particular reference to the methods set forth in the UIV Leaflets 505-1 and 506. In this regard, it should be noted that the accurate calculation of allowable profiles for a vehicle requires some input parameters, eg, the flexibility coefficient that can be stored in the vehicle database for each known vehicle model and that of the measurements performed within the SMI can be difficult to calculate. The loading profiles defined by means of standard markings on the vehicles for combined transport and for the transport of special loads, as described above, are stored in the system in one of the storage means defined above and, depending on the result of the OCR or OCR similar read operation of such marks retrieved by the system.

The criterion for considering a single detected position M as a profile violation of the vehicle or the freight width may be defined by the following condition y m - y p (x m , z m ) - ε (κ εy , σ my )> 0, (126) where y _m , x _m, and z _{m are} the coordinates of the measured 3DD position, and y _{p is} the corresponding lateral Profile position and ε a tolerance _allowance , which is a function of a tolerance _factor κ _εy and the measurement uncertainty σ _my . The uncertainty σ _my clearly depends on the measuring instrument used for the definition of the 3DD position and the M coordinates. Ideally, σ _{my should} also take into account the uncertainty of y _p , which in turn depends on the uncertainty of Ω as the result of the measurement data used in the calculation and the margin of error in calculations. It is particularly evident that the error in the LFD, unless it has been reduced in the calculation of Ω, will have a direct consequence for the tolerance margin of ε, to an extent approximately proportional to the absolute value of the derivative of y _{p is} opposite z _m . Various mathematical expressions can be used for ε, but in all ε has a positive monotonic derivative _dependence of κ _εy , a value that can be "tuned" to the ratio between the frequency of missed detection of dangerous elements and the frequency of false ones Recognize recognitions even further.

The expression x m - x p (y m , z m ) - ε (κ εx , σ mx )> 0, (127) where x _{p is} the vertical coordinate of the relevant profile, κ _εx the corresponding tolerance _factor and σ _mx the uncertainty to be applied, may be used for the condition of profile violation of horizontal or near horizontal segments of the upper profile part.

Another term ξ (actually ξ _y or ξ _x ) can be added to the left expressions of 126 and 127, especially in the case where the corresponding profile is the allowed profile according to the principles of UIC 505-5 [052] , what about the conditions y m - y p (x m , z m ) - ε (κ εy , σ my ) + ξ y > 0 (128) and x m - x p (y m , z m ) - ε (κ εx , σ mx ) + ξ x > 0, (129) which forms the basis for a set of infrastructure and train operating conditions, with particular reference to the comparison of the coordinates M with the boundary profile of the vehicle design. A component ξ _i of ξ can be defined by considering the actual boundary profile of obstacles along the rail to which the hazard detection relates. The value ξ _i generally represents a function of the vertical coordinate in the plane of the normal coordinates and it could be different for the two halves of such a transverse section of the plane for the opposite sides of the vertical axis. In fact, as a positive component, _i considers an additional breadth of the infrastructure profile and, as a negative component, the presence of particular infrastructure profile constraints. The recent availability of special gauges or vehicles that can detect the position of obstacles on the rail allows railway companies to determine updated and reliable values of ξ _i that can be stored in the system. A second component ξ _k of ξ can be defined taking into account the effect of the maximum vehicle speed. The use of the term ξ _i can be very valuable for managing alarms with respect to boundary line profiles and the generation of warnings regarding the transit of vehicles, eg the transport of exceptional cargo, which can be allowed if the train speed is reduced accordingly, so that the values are suitable for the whole or part of the way. Depending on the type, performance, and installation geometry of the instruments, the generation of alarm limits for boundary line profiles and of warnings can be conditioned by the detection of two or more points M, which form a sort of clustering. In such a case, expressions 126 and 127 (or 128 and 129) may be used to select one of these items, and different conditions may be defined, taking into account all elements of such accumulation to reduce false alarms. A particular condition that can be used to reduce the rate of false alarms, depending on the suitability of the 3DD measuring systems, is the presence of near intermediate points M from the outer impermissible positions of M, towards a set of points M within the allowable space, so that you can exclude small flying objects such as single sheets or small paper clips.

One of the most important means for reducing false alarms is the requirement that an alarm be consistently generated by processing 3DD data for one or more positions along the route, such as suggested at 004, for example. The criterion of consequence for the subsequent recognitions typically requires that in the case of this invention the possible hazardous or abnor male element has moved almost integrally with the vehicle. This alarm filtering technique can add considerable additional cost to the system hardware if its implementation means installing on the SMI additional expensive meters, but this need not be the case if the installation of such instruments is in any case facilitated by the use of their measurements to calculate the Ω parameter is justified. In addition, it is generally possible for the subsequent detections based on the 3DD measurements to be performed by other types of instruments. This technique is very effective in suppressing a significant fraction of possible false alarms for gauge lines and can be easily implemented with little developmental change in the corresponding software. One particular group of dangerous deviations from the gauge line comprises loose or torn wagon panels, covering or wrapping sheets of single loads on an open wagon, as in these cases the detected accumulations of 3DD points can flow significantly between two successive detections in the SMI. Although it is possible to distinguish some alternative methods with good statistical performance in detecting these cases and distinguishing them from flying shopping bags or plastic strips, Applicant recommends that such a distinction be made by remote control, such as is explained in more detail below. The detection of hazards of the contour line profile of a particular rail vehicle requires that the appropriate set of 3DD measurements be retrieved from the entire set of 3DD measurements. Such a data retrieval process may be performed by calculating a measurement time range for each 3DD measuring instrument, taking into account its installation geometry, calibration parameters, the Ω-coordinate (or LDF) transform member, and a small margin, ideally as a Function of the vehicle speed is calculated when passing the SMI. Although the entire process of comparing the 3DD points may take place in coordinate space C _VB , it may be advantageous to first select a selection of those 3DD points that might correspond to the dangerous element in coordinate space C _GB using a "conservative" profile to process.

The admissibility a particular vehicle and cargo boundary profile according to the principles the standard UIC 505-5 hangs along with the actual or assumed profile obstacle along the route and with the speed of the train. To the usefulness of the system for the largest possible number of installations, you should therefore design the software so and develop that they are the diagnostics of gauge line profilers for one Set of different alarm conditions (especially for conditions 128 and 129) perform effectively different infrastructure outlines and different ones Cruising speeds of the train.

The above-described techniques for detecting dangerous Deviations from the boundary line profile are available for many possible use depending on which profile definition is used and how the system information and notes with specific information system the railway company or manned control centers. Some possibilities be here on incomplete Way explained below, in connection with the generation and administration of diagnostic information, Alarms and data for Boundary line profiles.

The Detection by the system according to the principles UIC Leaflet 505-5 [052] "Incompatibility" of vehicle profile, including cargo, if available, and the infrastructure profile, taking into account the value ξ, if appropriate and possible, should lead to the generation of a boundary profile alarm, the sent directly or indirectly to the signaling system of the railway company should be to stop the train at the first possible position or to redirect him to a safe haven.

An alert should be sent, directly or indirectly, to the signaling system of the railway company (to stop the train at the first possible position or to redirect it to a safe harbor), even if the violation of a loading profile of the following types has been detected:
loading profiles defined by the RIV Agreement [060] for loads loaded on normal flat and open wagons;
load profiles defined by the RIV Agreement [060] and the UIC Code Leaflets for combined transport [054, 055, 056];
load profiles defined by RIV Agreement [060] and UIC Code Journal 502 [075] for the transport of special loads (special deliveries) using encoded profiles.

Such detected loading profile injuries are indeed likely associated with the presence of a shifted load or improperly loaded cargo. Of course, the recognition of a specific loading profile by the coded mark on a vehicle or its cargo in Verbin in that no violation of the cargo is detected, the system will be used to generate a possible alarm or warning regarding loading profile violation as defined by the RIV Agreement [060] for loads on a normal and flat wagon, to suppress. In this way, the system prevents the generation of a number of false alarms and warnings for special deliveries.

in the Generally one can have two or more tolerance limits for comparison define 3DD data with the profiles and a warning (plus data & information) instead of triggering an alarm if stricter tolerance values to meet an injury (during the looser values do this). A balance between sensitivity and the rate of false alarms can be achieved in this way so that the hazard reduction level is preserved with the use system without compromising rail traffic.

In all cases a profile violation (maximum permissible profiles, loading profiles and construction profiles) it is possible and recommended to use a Send warning to a control center, along with data and information that a remote technician and / or software application possible to assess the case and the appropriate measures to take (for example, an alarm to a signaling and security system send, send a message to the relevant platoon commander / crew, or a system-generated alarm that is already on the signaling and security system was sent to suppress). It may also be that the generation of an alarm and / or of Warnings are suppressed by the system when the system of a railway system has received the information that a particular abnormal situation related to the gauge line profile (i.e., the corresponding vehicle will be intentionally over Loaded standard boundary line profile). If an information record from the system for the train has been received, including the railway a series of sections with specific infrastructure profiles and / or reduced speed planning is recommended that the system takes this information into account so that the detection functions Correctly performed in connection with dangers of the gauge line profile. It may also be that the system the information system of the railway company sends a record to the gauge line profile for all the vehicles, for the a violation of their load limit line profile or an injury a certain maximum boundary line profile was detected (e.g., the ones corresponding to the standard datum line profile of leaflet UIC505-1), and that the information system of the railway company these records used to be right now or later to perform a compatibility check, depending after the actual Track of the train. In this case, the record profile data for the Vehicle and its load included, or it may consist of a compatibility mark for one Set of infrastructure profiles known to the system exist, if possible in the form of a function of vehicle speed.

A number of specific diagnostic methods can also be implemented to detect violations of RIV or specific loading rules that may not necessarily be classified as related to the perimeter line profile. An example of such an injury is insufficient longitudinal spacing between two loads on two different side-by-side rail vehicles, with the first load projecting beyond the vehicle on which it is loaded to the second vehicle, ie, with respect to the charging instruction 4.3 4, Volume I, Annex II of the RIV Agreement [060]. Such methods would have in common with the boundary line profile diagnostics discussed in this section that they relate to a measurement of a point, a zero, and a direction of the vehicle based on information regarding the rail vehicle model and the application of Ω or Ω ^-1 links.

Various Types and models of rail vehicles have a different one size Probability of dangers of gauge lines on, i.e. in connection with the possibility of improper loading or load shifting, and it recommends therefore, the stronger endangered vehicles before the less endangered to analyze the average time for the display of important Reduce the dangers of the gauge line profile.

While the system may apply the boundary line profile diagnostics to any rail vehicle, it may be decided not to test certain whole groups of identified rail vehicles, especially passenger cars. However, it is recommended to perform a basic compatibility check of the gauge line for all vehicles, and to check if their model is approved for use on a line marked by the corresponding gauge line profile. The basic control can be useful especially if the vehicle database has not yet been completed and if the full data sets of the vehicles, where the risk of danger of the gauge line profile is greater, are supplied and / or coded with a higher priority.

5.10 detection of dangerous Deviations from the boundary line profile for the lower parts of an identified vehicle

The Detection of dangerous Deviations from the boundary line profile for the lower parts of an identified Vehicle is treated here below, separate from the previous one Discussion about the Car body, in accordance with the separation of the discussion for the low and high parts of vehicles according to the leaflets UIC 505. Also is this separate discussion the profile diagnostics of the lower vehicle parts justified because of the different danger and the different effects in terms of measuring instruments, in need of installation at the SMI.

Also if the 3DD gauges for the lower vehicle parts taking into account the same principles like the 3DD measurements for the upper parts are chosen and can be installed stand up for the lower parts have some extra Difficulties. Such instruments are to be installed near or under the rolling surface be, and their frontal appearance or the windows must be horizontal or at a greater angle of inclination be aligned, which is related to the known problems with pollution, projected pebbles, weather influences strong air turbulence, fat and so on. In addition, the 3DD gauges should be compatible with the rail vibrations, or with a suitable decoupling be installed, and they could special maintenance or reinstallation during route maintenance need. The applicant holds However, these difficulties do not pose a problem that is innovative solutions requires, since even in the worst possible positioning the Professionals of this technique apply other alternative technical solutions can, the previously for the "hot box detectors" were applied (e.g., open protective cover for measurements, housing in hollow sleepers, clamping systems for the Mounting on rails, vibration damping for sensitive components, etc.). This possible Difficulties in development, installation and operation can However, to contribute to the decision, the system without the measuring instruments to implement that for the detection of the dangers of the gauge line profile of lower Vehicle parts needed become.

If The functions implemented by the system include the detection of dangerous ones Deviations from the boundary line profile of lower vehicle parts includes, then the calculation methods for this are basically the same as those, the above for the upper vehicle parts have been described. The detection of dangerous Deviations from the boundary line profile of lower vehicle parts In addition, the aspects would from leaflet UIC 505, the admissibility of vehicles on runoff mountains at free-fall yards.

5.11 Thermal diagnosis for axle bearings, Wheels and brakes

The following text deals with the functions, according to Box 239 from 3 to detect a range of abnormal conditions and axle box, wheel and brake failures that are based on passive infrared sensing for the vehicles whose model has been positively identified.

5.11.1 Sensors for the thermal Diagnosis for Axle bearings, wheels and brakes

In accordance with the data processing method mentioned below, as well as with the possible changes in detail, more than one type of infrared sensing device or more than a combination of devices meets the minimum space coverage requirements, the immediate field of view for a single pixel or sensor , the accuracy and the measurement rate. Thus, below are some alternative options of the apparats or apparatus groups (here called "BWBTIS" for Bearings, Wheels and Brakes Thermal Infrared Sensors) that can be integrated into the system to make measurements that allow the system software to cope with a number of abnormal conditions The Applicant emphasizes that this overview is deliberately brief, as many of these questions are the subject of a number of previous patent documents, especially IPC-B61K9 / 06, as well as the patent documents [005, 008, 016, 018, 019, 026] and some of the patent documents cited therein A first group of BWBTIS uses a single infrared detector or a few detectors with appropriate optical, electronic and mechanical equipment, thanks to which they are close to or directly on rails can be mounted on the rails or in hollow sleepers.The use of different types of sensor elements In particular, thermistor bolometers, LiTaO ₃ pyroelectric sensors, PbS and PbSe photoresistive detectors, HgCdTe (MCT) and InSb-Pho have been discussed tonendetektoren. Various technical solutions have been implemented in these devices to compensate for the temperature changes, to recalibrate the heat radiation measurement and, in some cases, to provide a signal relating to the difference between the temperature of the observed target and the ambient temperature. Many of the commercially available devices in this group are equipped with automatic protective covers for the optics to protect them when standing still, as well as with snow and ice heaters. A significant part of the electronics of these systems and the inventions in this particular field are related to the analog and / or digital processing of signals from infrared detectors, as this increases the sensitivity and lowers the rate of false alarms. However, the system mostly depends on such subordinate electronics or data processing methods. The electronics for signal processing such products are generally mounted with the sensors, or even in separate units for signal evaluation and processing (in this case, the sensor device is generally called "HDB scanner"), while the signal evaluation and processing unit is often " Detector is called). Some of the commercial devices of this group can be connected to the system in various ways, possibly with some modifications, the only fundamental requirement being that they retrieve the signals from the thermal detectors with sufficient resolution, measurement rate and accurate timing that the system software associate each measurement with a time that can be accurately related to the times at which measurements were taken from the other sensors and instruments installed at the SMI. Other signals from these devices, such as for calibration and diagnostics, may be provided by the system, if useful and appropriate. The control of the BWBTIS devices of this group, including, for example, the opening and closing of the protective covers, the temperature control and the calibration hatches, can be partly transferred to the system software via a suitable hardware interface, or it is performed only by the electronics of the devices. The choice of the extent of possible changes to these devices to integrate them into the system generally depends on economic considerations that take into account development, industrialization, production, and maintenance. An example of a commercial BWBTIS of this group is the Sentry System Sampling Unit of Southern Technologies Corporation [962], which is used for aiming on axle boxes or wheels, depending on the installation means. General Electric Transportation Systems [963] offers a range of BWBTIS devices of this first type that can be integrated into the system, including the Advanced Concept Scanner (ACS), the Vertical Look Scanner (VLS). , the "HWD"("Hot Wheel Detector") and the "FUS"("High End Hot Box Detection").

A second group of BWBTIS devices, the for Integration into the system is made up of one or few fast infrared sensors, i. Photon detectors, with one Mirror scanning system that the sensor beam in a plane, or near to a plane that is orthogonal to the direction of movement of the Vehicle is located. equipment of this type ("VAE-HOA / FOA 400 ") are from UAE Railway Systems [964] available in a few different versions, and they also offer different mounting options. Depending on the installation of the scanner, monitor this scan detectors the axle bearings from both sides of the wheels, the axle itself, the wheels and the brake discs, if any.

A third group of BWBTIS corresponds to the linear infrared imaging devices based on linear arrays of infrared sensors. These image-forming devices belong to the group of "staring arrays" (a term widely used in relevant publications), as opposed to the fast photon-sensing devices, which produce devices mounted with a vertical or near-vertical viewing plane a series of linear images, with the advantage of more continuous or near-continuous spatial coverage over the devices of the first group, and relatively small sensor points on the measured targets corresponding to each pixel of the array. 17a and 16b show two idealized views of a linear infrared sensor close to a rail 655 or 658 for making measurements of heat emissions at different wheel parts, bearings and brakes, which become one axle 645 or 633 belong. The viewing plane of the linear infrared imaging device 650 or 638 is vertical, or almost vertical, even if it can be tilted, if this is more favorable. The imaging device has a field of view that the angle 647 between the viewing directions 660 and 648 corresponds to the two extreme pixel values of the linear sensor array used. The central view direction 652 divide the angle 647 into two equal parts and defines a relative angle of rise 651 opposite the rolling surface plane 649 , The observation of such elements as brake discs requires a sufficiently oblique angle 635 between the viewing plane and the normal one 636 to the rail axis. The optical measuring beam 637 falls with the beam 646 together, while the lines 639 . 640 and 641 Plotting planes or rays at different time points over time, for which the longitudinal position of the axis of the measuring beam 637 equivalent. A measurement like the one through the beam 646 or 637 corresponds to an angle 653 depending on the corresponding pixel value and positioning of the image forming device, and a time value. Such an angle could of course be for one other viewing direction are measured, such as the center beam 652 , The ability to make measurements of different elements is clearly dependent on the position of the imaging device, its viewing angle 647 , the rise and fall angles 651 and 635 as well as the possible inclination of the viewing plane with respect to the rolling surface. Generally, at least two imaging devices are required to observe both wheels of one axle and all the brake discs. The positioning on each rail side of two imaging devices with opposite angles 635 could offer some advantages in terms of visibility of the targets as well as temperature differences between the front and back of such elements as the bearings, but is not a general requirement. The longitudinal gap 642 between two consecutive measurements for the same pixel value for the linear array of infrared sensors for viewing adjacent beams or planes 640 and 641 is, of course, of considerable value, especially in the context of detecting overheating of relatively small elements such as the bearings 630 and 654 , Depending on the particular type of linear infrared imaging devices, the pixels of the linear array perform their measurements synchronously or sequentially. It is thus also possible to arrange the pixels in two or more parallel lines instead of along a straight line and in this case they are associated with different skew angles to be taken into account in the data transfer and the processing hardware and / or software.

Very few options are available in practice for the choice of the linear array of infrared sensors, since a relatively fast response time due to vehicle speed is required and also because the use of expensive sensors and / or low temperature cooling systems has two important negative elements in system implementation are. In fact, while the number of pixels required for this application to achieve the desired spatial resolution depends on the viewing distance and the visual field of vision, the longitudinal resolution (ie in the direction of travel of the vehicle) is limited by the response time of the sensors of the array. A vehicle with a speed of 120 km / h shifts in the longitudinal direction (distance 642 from 17b ) is about 33 mm per millisecond (or 22 mm per millisecond at 80 mh) and it is thus clear that the appropriate sensors should have a response time that is at least in the range of 10 ^-3 s to blurry images in the direction of motion avoid and allow an adequate longitudinal measurement gap, as mentioned above. The required short reaction time excludes the most popular and readily available types of uncooled staring arrays, namely the thermistor bolometer matrices, whose reaction time [068] is much too long, as explained by Yaktine et al. [026] There are two types of fast infrared arrays that are particularly attractive for this application, namely the "microthermopile" arrays and the photo-line sensor arrays, especially those for PbSe, because the lower wavelength sensitivity of PbS sensors (about 1 to 3 μm about 2 to 5 μm for PbSe, depending on the sensor temperature) makes it less suitable for the temperature range of the target.

Honeywell Inc. (Plymouth, MN, USA) has developed a technology for the fabrication of microthermopile arrays, or "thermoelectric arrays," a silicon monolithic structure with nitrided silicone bridges that overhangs the hot thermoelectric terminals support the micro-wells in the silicone underlayer [068] with the corresponding cold junctions near the edge of the micro-wells. A "speed optimized" 96 pixel array of this type with 150 by 150 μm sensors was used by ISI (Plymouth, MN, USA) [068] to develop the "IR 1000" model, a high speed radiation meter for imaging. The "Model IR 1000" pixel was designed [068] for a thermal response time of 7.5 10 ^-4 s, so that a high speed operation is achieved by sampling the signals for 8 10 ^-4 s, and an array sampling rate of 10 ³ s ^-1 , The spectral range of the sensor (8-12 μm) is particularly suitable for performing thermographic measurements at relatively low ambient temperatures and for reducing sensitivity to reflected and diffused solar radiation. NEID (Noise Equivalent Temperature Difference) for a f / 0.8 germanium lens is about 0.7 K for room temperature targets and the measurement range is from about 30 to over 400 degrees Celsius with a short-term accuracy of about ± 4 degrees Celsius or 2%. The two main advantages of this type of instrument in this application are the room temperature operations without any type of cooling device for the sensor assembly, as well as a consistently good stability. The fact that no breaker is required reduces costs and maintenance. In addition, the absence of a breaker allows implementation of an internal or external sampling trigger. An excellent description of the characteristics of this device and its use for adopting thermal imaging of wheels of rail vehicles and axle boxes can be found in a patent document [026] by Yaktine et al. from SAIC (Science Applications International Corporation, San Diego, CA, USA).

Photoconductive PbSe linear matrices provide sufficient infrared response for this application In the thermoelectric cooling temperatures in the 2-5 micron wavelength band with a bandwidth that is well above 10 ⁴ s ^-1 . Several different manufacturers supply pre-packaged PbSe matrices, with or without multiplexing devices and amplifiers, that can be integrated to provide a linear infrared imager. An example of such a product is the M-2105 series from Northrop Grumman Electro-Optics Systems of Tempe, AZ, USA, including a 128-element die in line with 91 by 102 μm pixels and a 256-element bi-linear die in time staggered 38 by 56 μm pixels. Although photoreactive detectors can be read in DC or AC mode, with or without a quick comparison to the reference target, the very high dependence of pixel resistance on pixel temperature for quantitative measurement applications requires the use of choppers, which typically accommodate the sample rate of the device about 2 10 ³ s ^-1 , which is however compatible with the present application. NETD values below 1 K can be achieved for target temperatures of more than a few tenths of a degree Celsius using f / 1.0 or better silicone lenses. Controlling and / or measuring the temperature of the breaker and other parts of the imaging device that affect the measurement makes it possible to achieve adequate accuracy over time for this application. These sensor arrangements therefore offer an alternative to the thermoelectric arrangements discussed above and offer the important advantages of high pixel count and fast response time, but also the significant disadvantage of requiring in practice a circuit breaker and a thermoelectric cooler. However, other types of linear infrared sensor arrays can not be excluded for this application, such as the MCT arrays used in the design described in Patent Document [019] or the few LiTaO ₃ arrays disclosed in the patent document [018] are described.

One can also use thermal FPA (Focal Plane Arrays) imaging devices, as explained for example in some earlier patent documents [003, 004], and they form a fourth BWBTIS group. However, the image processing speed required in this application to avoid fuzzy longitudinal images restricts the choice of commercially available FPA thermal imaging devices to a subset of products that offer some advantages, for example with respect to NEID, but are generally more expensive than the fast linear imaging devices discussed above. In addition, in this application, the FPA imagers would be used at a frame rate which, when appropriate, would not exceed about 10 ² s ^-1 because higher rates would result in a further increase in cost and because most of their pixel data would be unnecessary , Thus, assuming that a vehicle shifts approximately 330 mm in 10 ^-2 s at a speed of 120 km / h, this results in a significant difference in viewing angle for individual targets, with possible negative performance and performance effects data processing complexity. Some FPA imaging devices are available which allow for fast "windowing" of a pixel subset, but their advantage over cheaper and simpler linear imaging devices is at least questionable in this application. Another advantage of linear imaging devices over FPA infrared imaging devices is the easier protection against weathering, dirt, dust, projected pebbles, etc. Applicant is therefore generally not in favor of the use of FPA thermographic imaging devices in this application.

Applicant clearly states that various combinations of passive infrared sensing devices from the same BWBTIS group or from different groups may be considered for system implementation. For the reasons explained below, the BWBTIS gauges should be installed as close as possible to at least one wheel sensor, and preferably to a pair of wheel sensors mounted at a close longitudinal distance from each other on the rails or to more than one such pair. In addition, provided one or more fast precision distance measuring devices as described in 8th should also be mounted longitudinally as close as possible to the BWBTIS gauges, for the reasons explained in the next section of the text.

5.11.2 Data processing for the Thermal diagnosis of axle bearings, wheels and brakes

All measurements made by any BWBTIS consist of a scalar value that approximates (or can be converted to) the surface temperature of the observed element, with a temperature accuracy that depends on the instrument calibration, the long-term measurement stability, and the measurement stability in relation to the ambient temperature and heat radiation, the presence of interference and infrared emission properties (eg, emissivity to wavelength and temperature) of the corresponding surface point depends. In addition, if a measuring beam, which is defined by a sensor element and by the optics, does not hit a homogeneous surface, but for example on two different surfaces (this for example in the case of measurements on a Area boundary), or if the surface has a temperature gradient, the obviously measured temperature value is an intermediate value of the corresponding maximum and minimum values, the higher temperatures usually weighing more since there is no linearity of the dependence of radiated energy on the temperature. Each measurement is associated with a time, with a directional vector and with a position from which the measuring beam is directed at the target. The time will be either the system time or a time that is precisely tuned to it. The direction and zero point of the measurement beam are defined in the calibration, and possibly through the coordinate conversions in the coordinate system C _GB as defined above, or in other coordinate systems based on the ground, as discussed above in connection with other types of system measurements. A BWBTIS measuring instrument is characterized by the divergence of the measuring beam or by beams that are non-circular and can be defined in practice by an elliptical cross section with respect to the measuring distance. Moreover, the time interval for which the measuring instrument integrates a sensor signal is important information used in the corresponding data processing methods. As described below, the system uses measurements of heat emissions to diagnose abnormal and / or hazardous conditions of bearings, wheels, and brakes based on the information from the vehicle database regarding the corresponding identified vehicle model and the exact location information of the wheelsets in space as a function of time. From the text below, it is clear that the accuracy in associating a measurement point on an element surface depends on the accuracy with which the direction and the zero point of the beam are known, as well as on the timing accuracy and the accuracy in the assignment of a position over time considered element.

A first step in the diagnostics procedure for elements mounted on the axes is to define the time-dependent coordinate transformation function Γ _{WS used} in associating BWBTIS measurements with such elements, similar to the conversion function Ω described above a vehicle box. 18a and 18b show two views of an axis 683 or 696 over the rails 671 and 672 or 691 and 693 , similar to 17a and 17b , The same ground-based coordinate systems C _GB can, similarly as in 14 or in 16 , with the axis zero at 682 or 697 be used. The axis X _GB or 681 , in the 18a , pointing up, is invisible and runs in 18b beyond the leaf. The axis Z _GB or 687 , in the. 18b pointing along the rails is invisible and runs in 18a beyond the leaf. The axis Y _GB or 677 or 684 is visible and lies in the leaf level both in 18a as well as in 18b , A new coordinate system C _WS is introduced here to define the positions of the axis-related elements. The position 679 or 685 the zero point of C _WS in 18a and in 18b ideally lies on the vertical axis of the corresponding bogie box at a height that can refer to the corresponding axle (a possible alternative position is suggested below). The X _WS axis 678 (not visible in 18b ) is orthogonal to the axis of the axes and the rails. X _WS will generally not be exactly parallel to X _GB , as there may be an exact orthogonality between X _GB and the rolling surface, or because of the small roll angle of the conical shape of the wheel tires (not visible in FIG 18a and 18b ) and from the lateral shift of the wheel group. The Y _WS axis 675 or 688 is nearly parallel to the rolling surface (due to the said effect of the tapered shape of the tires) and is generally not closely parallel to Y _GB because of the variable yaw of the bogie or the rolling motion Ψ. The Z _WS axis 680 or 689 is practically parallel to a line passing through the middle of the two outer axles of a bogie and is generally not closely parallel to Z _GB because of the variable yaw of the bogie or the rolling motion Ψ.

In accordance with the methods described below for the diagnostic procedure for components mounted on the axles, the applicant does not require rigorous accuracy of coordinates conversion by the Γ _WS . However, some options for obtaining Γ _WS are defined below and, together with the uncertainty of certain measurements, affect that accuracy. However, it is clear from the text below that a more accurate conversion of Γ _WS reduces the margin of uncertainty in the application of the diagnostic criteria, and that such a reduction in uncertainty in spatial resolution and in a short reaction time of the BWBTIS instruments is more pronounced.

einer kombinierten Rotations-Übertragungsumwandlung in homogene Koordinaten durch die Umwandlungsmatrize [ΓWS(t)] = [ΓWSRZ][ΓWSRX(t)][ΓWSLD(t)] (131)in der die Zeitabhängigkeit praktisch auf die Übertragung und Yaw-Rotation beschränkt werden kann, und der Rollwechsel vernachlässigt wird, der beim Ändern der lateralen Position der Radgruppe im kurzen Zeitintervall stattfindet, in der die Funktion Γ_WS auf eine bestimmte Radgruppe angewandt wird. Der Anmelder stellt klar, dass die Roll-Abhängigkeit von der Zeit jedoch bei offensichtlichen Änderungen der mathematischen Formeln und der unten beschriebenen Methode ggf. mit berücksichtig werden muss. Der Wert der Winkeldrehung des Zwischenraums wurde hier weggelassen, da er in diesem Fall praktisch vernachlässigt werden kann.When using the RPY (Roll-Pitch-Yaw) conversion convention, the function Γ _WS in the matrix can be expressed by the following formula

a combined rotation transfer conversion into homogeneous coordinates through the transformation matrix [Γ WS (t)] = [Γ WSRZ ] [Γ WSRX (T)] [Γ WSLD (t)] (131) in which the time dependency can be practically limited to the transmission and yaw rotation, and the roll change that occurs when changing the lateral position of the wheel group in the short time interval in which the function Γ _{WS is} applied to a certain wheel group is neglected. The Applicant makes it clear that, however, the roll dependency on time may need to be taken into account in the case of obvious changes in the mathematical formulas and the method described below. The value of the angular rotation of the gap has been omitted here, since it can be practically neglected in this case.

wobei X_WS(t), Y_WS(t) und Z_WS(t) die Übertragungskomponenten von C_WS gegenüber C_GB, sind, Θ der Rollwinkel und Ψ(t) der zeitabhängige Yaw-Winkel.Thus, the transformation matrix for the transmission of the two relevant rotations looks like this

where X _WS (t), Y _WS (t) and Z _WS (t) are the transfer components of C _WS to C _GB , Θ the roll angle and Ψ (t) the time-dependent Yaw angles.

The inverse conversion of vectors from the wheel-set based coordinate system C _WS to the ground-based coordinate system C _GB can be performed by the following formula V GB = Γ -1 WS (T) V WS , (135) in which Γ -1 / WS expresses the inverse of the function Γ _WS , for which the corresponding template for homogeneous coordinates is defined as follows [Γ -1 WS (t)] = [Γ -1 WSLD (T)] [Γ -1 WSRX (T)] [Γ -1 WSRZ ] (136)

The definition of Γ _WS thus requires the assignment of a value to the parameter Θ and the definition of the time-dependent functions _XWS (t), _YWS (t), Z _WS (t) and Ψ (t). The relatively small dimensions of the axis along with the components mounted on the axle and the short longitudinal distance between them and the sensors whose measurements are used to calculate the time-dependent components of Γ _WS , makes it very easy to express X _WS (t) , Y _WS (t), Z _WS (t) and Ψ (t) for the short time interval corresponding to the application of the function of Γ _WS for a particular wheel group. First of all, the linear dependence might be sufficient to express X _WS (t), while parabolic expression would satisfy the expression of Y _WS (t), Z _WS (t), and Ψ (t). However, the discussion below on the calculation of the function Γ _WS shows that the quality and quantity of measurements used to define X _WS (t), Y _WS (t), Z _WS (t), and Ψ (t) may be such that some or all of these time-dependent amounts may be considered to be constant or linear.

An alternative special position of the zero point of the axes C _WS is the center of the axis, ie the axis of symmetry point of the axis which is equidistant from the wheel flanges. , However, one should consider regardless of the choice of the position of the origin of the axes C _WS, the axis as a circular symmetrical element, with the exception of the camp, which is not usually symmetrical and does not rotate. Consequently, the rotation of the axle and the components attached thereto can generally be neglected, with some possible exceptions associated with, for example, the wheel body of corrugated wheels or with slots and holes in some visible components of the brake disks. In other words, the C _WS coordinate system is integral with the axis of an axle and the axle-related components, such as the bearing shells, which do not roll over the rails. In view of these considerations, the gap value in the above Expression 131 was neglected.

in der die U-Werte δ_u und ihre entsprechenden Unsicherheiten σ_u den übereinstimmenden Positionen der Radsatzelemente im Zeitverlauf bei einer oder mehreren Messungen entsprechen.The calculation of the parameters defining the corresponding angular rotations and the linear displacement of the components of Γ _WS can be done by minimizing the quantity δ ² , which can be expressed by the following chi-square formula

in which the U values δ _u and their corresponding uncertainties σ _u correspond to the coincident positions of the wheelset elements over time in one or more measurements.

Assuming that the wheel sensors are precise enough, then the times t _{j ', k} relating to passing the center of an individual wheel j' at the individual sensor k 'can be used directly for the determination of the components Ψ and Y _WS of Γ _WS may be useful by the following expression δ u ' = z WS (j ', t j ', k' ) - z WS (k ', t j ', k' ) (138) where z _WS (j ', t _{j', k '} ) is the z-coordinate of the corresponding wheel center in the coordinate system C _WS and z _WS (k', t _{j ', k'} ) is the z-coordinate of the corresponding wheel center in the coordinate system C _WS .

in der die Integrationsgrenzen in Bezug auf ein Messintervall für den Radsatz unter Berücksichtigung der Geschwindigkeit gesetzt sind, können verwendet werden, um die erste und zweite Ableitung im Zeitverlauf in Übereinstimmung mit der LDF-Funktion zur Definition von Z(t) zu bestimmen, und somit können andere Werte wie 138 zur Definition des Offset verwendet werden. Diese Bedingungen können natürlich auch anders definiert oder geschrieben und, je nach den Interpolierungsexpressionen für Z_WS(t) und für Z(t), in geschlossene Formelgleichungen gesetzt werden. In der Tat besteht im Allgemeinen ein Offset zwischen der LDF oder Z(t) für ein Fahrzeug und Z_WS(t), doch dieses Offset ist in den Expressionen 139 und 140 irrelevant, wo die erste und zweite Ableitung von Z(t) und Z_WS(t) im Zeitverlauf verwendet werden.The two terms δ _u

in which the integration limits are set with respect to the speed with respect to a measuring interval for the wheelset can be used to determine the first and second derivative over time in accordance with the LDF function defining Z (t), and thus For example, other values such as 138 may be used to define the offset. Of course, these conditions can also be defined or written differently and, depending on the interpolation expressions for Z _WS (t) and for Z (t), set in closed equation equations. In fact, there is generally an offset between the LDF or Z (t) for a vehicle and Z _WS (t), but this offset is irrelevant in expressions 139 and 140 where the first and second derivative of Z (t) and Z _WS (t) over time.

If one or more laser distance meters similar in 8th their data can be useful in determining Γ _WS . In particular, the use of the method discussed above for element F1 (in the calculation of Ω) with one or more fast and precise laser distance meters is very effective in determining Y _WS (t) and contributes to the definition of Ψ (t) , depending on the zero point selected for C _WS . The same data would also be useful in defining Z _WS (t) and Ψ (t) by detecting the front and rear wheel edges. The corresponding expressions and calculation algorithms for the values δ _u can be easily defined by taking into account the examples given above for the definitions of the alternatives of the term ζ _r .

X _WS (t) can be defined by measuring the wheel diameter (see discussion of 8th above), taking into account the actual position of the railheads, as explained further below in the discussion of the system calibration. In addition, if individual linear VIS or NIR image forming devices (or fashions of linear image forming devices with the viewing planes preferably located in almost the same vertical plane) are installed at a low height, so that high quality images (or high quality stereoscopic images) In addition, the round elements corresponding to the flange edge or the round edges on the outside of the wheel can be recognized and localized in space-time relationship. The location of the round elements can be particularly reliable because the information from the vehicle database on the (non-worn) wheel, along with the timing of the wheel sensors, greatly limit the allowable range of position, orientation and curvature. The relevant computation procedure may be to first select a series of pixels that match the allowable arcs in an image and, in a second step, seek the correct match of the actual round wheel elements. The corresponding δ _u expressions would look for the correspondence between fixed or variable diameter circles (depending on wear) and the imaging vectors for the selected pixels at the corresponding imaging times. Such a method of processing the linear mappings allows the definition of X _WS (t) Y _WS (t), Z _WS (t), Θ, Ψ (t) and the rolling radius with a very high accuracy.

you The bearing shell can also by pattern recognition and position matching Recognize by looking at the picture of the bearing shell and its position in relation to the center of the axle from the vehicle database. In In this case, the corresponding figures for the position match should be but with a very similar one Imaging geometry by the same or equivalent linear imaging devices be determined. Such a need for a great similarity of visual geometry however, can be reduced if the software component for pattern recognition relatively sophisticated or individualized algorithms used.

The values of σ _u for the expression 138 or for the value δ _u related to a fast laser Ab In principle, the figure-based method described above would, in principle, be described as a function of the accuracy of the sensor measurements, while for the expressions of δ _u, such as 139 and 140, an empirical choice would be preferable.

In general, the calculation of Γ _{WS is performed} by an iterative procedure for multiparameter search of the minimum expression 137. In the case of the method described above with linear imaging devices, the first convergence iterations should only consider the simplest and strongest terms of δ _u , such as 138, 139 and 140. A second step in the diagnostic procedure for axis mounted elements concerns the determination of the representative temperature values of corresponding elements or parts thereof. This is achieved by associating the BWBTIS measurements with the axis-related elements, knowing their surface position in the coordinate system C _WS from the vehicle database, and applying the Γ _WS coordinate transformation to the corresponding measurement vectors. The actual method must also consider the visibility of the elements, the uncertainty of the vector match, and the finite dimensions of the measurement beams.

The axis-related elements may be described in the vehicle database as a set of small polygonal flat surfaces (typically triangles or quadrilaterals as for the "3D mosaics") or a combination of polygonal surfaces, sets of circular surfaces, bevel frustum surfaces, circular surfaces, and so on. Each of these surface elements is associated with a mechanical element and a set of parameters that depend on the geometry and define its shape, size, position and orientation in the C _WS coordinate system. In general, it is advisable to set the entire "visible surface" on a mounted axle with its components in relation to these surface elements. Given that the same axles and wheelsets or even the same complete bogies are mounted on different rail vehicle models, it may be useful to store their surface description once and reference it in the vehicle database, taking the position of their C _WS - Indicates coordinate systems in the C _VB coordinate system of a vehicle model. In general, the degree of detail that should be used in this coding of axis-related surfaces depends on how critical certain geometric details are in terms of the corresponding performance and rate of false alarms of the diagnostic method or methods.

The specific parameters (i.e. the corresponding considered axis), which is in the procedure for BWBTIS measurements based on the diagnostic procedure to be used generally depending on the type of BWBTIS equipment used in the system will, and hang from the actual Processing method. In any case, they are in the vehicle database in association with an axis type or a lathe type or stored on a specific vehicle model, depending on the design of the Software implementation and the independence of such axis-related Diagnostic parameters of the actual Bogie or vehicle model.

Of the Term "HTDS" for "Homogeneous Thermal Diagnostics Surface " used here to make an area to define which one for may consider the diagnostic purposes homogeneous as seen in thermal terms. Some examples of HTDS are a vertical surface a bearing shell parallel to the direction of the pulling movement, a low part of a round surface of a bearing shell, an outer part a brake disc surface and a free round surface an axis. Another particular example of an HTDS is the lower part of a flat outer surface of a Radreifens, a significant because of a brake-locked wheel higher Temperature as the outer surface of the Wheel tire can have a total. An HTDS can be found in the vehicle database be encoded by one or more geometric surface elements. A handy way to encode a HTDS is to do this with a subset of geometric surface elements at the top the coding of axis-related surfaces have been mentioned. So the entire recognizable area can be a particular axis-related Elements of a set of coded surface elements put together being only one or a few of these particular HTDS for diagnostic purposes represent. A given geometric surface element can have more than one HTDS correspond, as in the case of different parts or the Total external area of one Radreifenebene.

A first method, referred to herein as "TAM1" for "Temperature Assignment Method 1", for assigning a representative temperature to HTDS applies to the case where one or more BWBTIS measurements can be safely related to an HTDS. For such TAM1 measurements, reading by the instruments should not be impressed by the temperature of any other nearby or adjacent HTDS surface in the field of the corresponding BWBTIS (without the actual HTDS temperature variation due to heat exchange with other surfaces to be considered). In practice, the uncertainty regarding the position of HTDS with respect to the Strahlstrahlvek and the measurement beam traces on the HTDS together so that BWBTIS sees only parts of the HTDS, ie no significant parts of any other surfaces. 19a is a conceptual illustration of the conditions under which TAM1 can be applied. The triangle 700 is an example of an HTDS while the two surfaces 701 and 702 behind and in front of the surface respectively 700 lie. The cross sections of the measuring beam of 703 to 708 ideally refer to a single passive infrared sensor and the entire image of 19a is defined as a view from an infinite distance along the optical axis of the measuring beam. The measuring beam point 703 is significantly larger than the others due to the divergence of the measuring beam. The larger ellipses like 709 and 710 give the range to which the corresponding measurement points could be referenced, taking into account the dot position uncertainty on the observed area. In general, the "geometric match uncertainty" varies depending on the direction, depending on the accuracy of the measurements and the Γ _WS transformation. The two axes of the measuring point ellipses and the larger ellipses, which result from the position uncertainty, usually do not match and in the drawing they are purely coincidental. In the example of 19a , only the measuring points correspond 705 and 706 strictly the suitability requirements for the application of the TAM1 on the 700 , The temperature assigned to the HTDS by the TAM1 is only the average of a single relevant temperature reading.

Obviously, the applicability of TAM1 through a more accurate Γ _WS conversion, more accurate HTDS definitions, larger HTDS dimensions, measurement viewing angles that are more normal than the observed surface, becomes higher data transmission rates, higher sensor bandwidth, narrower measurement beams, lower vehicle speed, closer Angular distances between the beams of two adjoining pixels (if applicable) and a faster mechanical sampling rate favors.

19b represents the same case as 19a but with a much greater uncertainty of measurement positioning along the series of measurement points 723 to 728 , In this case, the cases are like 729 and 730 for assigning a measurement to a surface element relatively wide, so that no measurement point certainly as purely representative of HTDS 720 can be considered as it partially with the other surfaces 721 and 722 could overlap. Therefore, TAM1 is not in the case of 19b applicable. A special feature of the measuring points and their respective position uncertainties in 19b is that at least two of the measuring points completely from the HTDS 720 must be covered. If this particular situation applies, a second method, called "TAM2" here, can be used to determine a representative temperature for an HTDS. TAM2 relies on the fact that the maximum difference between the temperature readings of these points completely from the HDTS 720 can be estimated by the Meter's Noise Equivalent Temperature Difference (NEID), as well as by the other contributions to the measurement spread that are based on the largest expected variability across the HDS of surface temperature and surface emissivity. Such a total scattering may alternatively also stem from the statistical processing of the actual measurements. TAM2 can be satisfactorily applied when the temperature readings for the "contiguous" surfaces such as 721 and 722 enough different from those of the corresponding HTDS. In fact, if these conditions are met, a search can be made based on a series of measurements within the abovementioned maximum variance and "surrounded" by statistically different values. The representative temperature for the corresponding HTDS according to TAM2 is thus defined as the average of the measurements found within the applicable deviation of suitability. 20a is another example of a possible measurement situation in which the width of the considered part of a HTDS is comparable to or smaller than the width of the measurement point. The area HTDS 741 is a HTDS, which is between two other surface elements 740 and 742 lies. The 14 measuring points as well 743 lie centered above the line 744 and mostly overlap with a few other points that are close to them. In this drawing, the areas are not specified that correspond to a position uncertainty. In this particular concept example, the pattern rate is such that the measurement points are largely one above the other. An example of a practical situation that is similar to this case is the observation of a part of a brake disk, as in 17a and 17b , by a sensor positioned outside the space between the rails and in which the effect of the divergence of the measuring beam causes the points to coincide with each other at the distance of the HTDS from the sensor. The 20b refers to the temperature measurements according to 20a , The axis 745 represents the actual or measured temperature, while the axis 746 a linear space coordinate above the line 744 can correspond. The actual temperature curve 747 assumes that each of the three surfaces 740 . 741 and 742 has a homogeneous temperature. Each of the measuring bars like 749 . 752 . 750 . 751 and 748 corresponds to reading one of the measuring points 20a , The vertical length of the measuring bars corresponds to the NEID of the temperature measuring device, and the clearly recognizable offset of the first and last measurements of the series, as well as 748 and 749 , opposite the curve 747 comes from a temperature measurement error related to the different emissivity values. A third model, here called "TAM3", can in certain cases like the in 19a be applied to assign a representative temperature to a HTDS based on the calculation of an approximate curve such as 749 in 20b , This type of problem of curve approximation is known from other applications and is discussed in many publications such as articles or book in the field of experimental physics. The actual shape of the curve depends on the beam profile and approaches a flattened angular peak as the spot size decreases in proportion to the width of HTDS. A precise definition of the function of the Γ _WS conversion together with an exact coding of the HTDS makes TAM3 particularly secure. In addition, if the measurement beam profile is also accurately known from distance, the approximation curve process may give an approximate estimate of the HTDS temperature, even if the ratio between HTDS width and dot width is smaller than in the case of 19a is, and, accordingly, a larger difference between the maximum measured value, such as 750 , and the actual HTDS temperature is. If the point positions are known with a high uncertainty (eg along 744 m in the case of 20a ), then the TAM3 method can be used, which gives the approximation function the necessary spatial offset flexibility, provided that it is known that the areas around the tip have a distinctly different temperature. The homogeneity of the temperature of the surrounding area is another factor that allows TAM3 to be used successfully in this last specific situation. TAM1, TAM2 and TAM3 are above with respect to FIGS 19a . 19b . 20a and 20b for a series of measurement points arranged along a row, but they should generally be considered applicable to measurement sets whose beams are scattered in two angles or in dislocation magnitudes, depending on the type of measuring instrument and the geometry used in recording the geometric data projection. A fourth particular method, here called "TAM4", may be used in certain cases to assign a representative temperature to one or more HTDSs for a data set comprising a few areal elements. A case in which this method is particularly recommendable is with a bearing shell together with the corresponding fixed wheel body, and the measurement of the temperatures from the rail side by a device of the second or third group of the above-mentioned BWBTIS. In this case, four HDTS could be defined for a representative portion of the cup surface, for two parts of the body on the front and rear sides of the rail side, and for part of the body just below the bearing. A temperature function of three space sizes in the coordinate space C _WS can be defined over the corresponding area (bearing shell and wheel center) with sufficient flexibility to describe the transition areas between the various HTDS, but with sufficient rigidity to allow reliable algorithm convergence A key benefit of this method is the ability to achieve good performance with a relatively low accuracy Γ _WS function, ie, with respect to the X _WS component This method can also be applied to the BWBTIS data series in a series, such as those generated by the first type of the BWBTIS mentioned above.

All four mapping methods discussed here require that series of BWBTIS data be recorded, with the Γ _WS transformation being defined by area elements in the vehicle database. As with the methods discussed above for the detection of boundary line profile errors above, the calculations for the correspondence of measurement beams and area elements in the coordinate spaces C _WS or C _GB can be made by taking the Γ _WS conversion over the measurement beam data or the inverse Conversion of surface element positions and orientations into ground-based coordinates used. The uncertainties of the relative positions and / or orientations of the measuring beam and the surface elements can be estimated using standard methods, depending on the data used to define the Γ _WS parameters and their accuracy and processing methods were used in the definition of the Γ _WS parameters. Various computationally effective algorithms may be used to define the intersection regions of the measuring beam and one or more surface elements, including some of those described in publications on the visualization of three-dimensional vector drawings or optical ray tracing. The conversion of all geometric measurement beam data from the C _WS to the C _GB coordinate system or vice versa is unnecessary and would mean a waste of computing resources. It is therefore recommended to apply the coordinate transformation only to the data of the measurements that are candidates for a possible match with the corresponding surface elements. The function Γ _WS can also be used for this by defining time intervals, and if applicable, pixels or scanning angle ranges, taking into account a margin of uncertainty (some bounding surfaces including axis-related components, and the BWBTIS data are suitable for performing the calculation ). Several algorithm solutions can be implemented by technicians to reduce the computational burden associated with processing the BWBTIS data. For example, several BWBTIS data can be recorded for a particular HTDS, and some subseries can be mapped without computing radiometric overlaps with the HTDS if they are geometrically secure between two or more measurements assigned to the HTDS. A fifth special method, here called "TAM5", for assigning a representative temperature to one or more HTDSs for a mono- or bi-dimensional dataset, if the possible variations with respect to the surface elements in question are relatively small (eg if an element is not hidden or due to a change in the bogie Yaw or due to different degrees Radabnutzungen not significantly smaller). TAM5 is based on the processing of a "pseudo-map" of axis-mounted components, ie, a monodimensional or bidimensional template of temperature measurements characterized by two spatial variables, eg, the viewing gap angle for the pixels of an infromastron template and the longitudinal displacement. The coordinates of the components of the pseudo-imaging die are slightly shifted or deformed to maximize the match based on the temperature of particular regions of one or more HTDS. An accurate value Γ _WS is probably not necessary in the case of TAM5. This method can be seen as a form of pattern matching that uses a pattern with some flexibility and allows some deformations of the record to match a fixed pattern. TAM5 differs from the above methods TAM1 and TAM4 because the three-dimensional coordinates are not explicitly used, but this method also uses specific information from the vehicle database. Other methods of temperature mapping may be defined by engineers experienced in the respective techniques, using geometric computations and signal processing methods to maximize the accuracy of the temperature estimation with respect to the HTDS elements defined in the vehicle database.

The for the Processing of the axis-related passive infrared sensor data to be used Methods are in the vehicle database along with all necessary parameters and information about their application.

The Applicant states that in the case of the integration of a commercial BWBTIS device or a modified version of such in the system, and if such a device produces a satisfactory estimate of the temperature of HTDS (absolute or environmental) caused by such Instrument determined value in the next step of the procedure for the thermal diagnosis of axis-related elements can be used. This last remark is also valid in the case that such integration consists in the use of a "scanner" and in that the corresponding data processing software runs on separate hardware which may be part of the system's calculation software. In these cases, the system is in any case an advantage over the prior art, even if the Γ _WS conversion is not explicitly used. In fact, the specific information from the vehicle database allows for a better balance between the detection rate of defects and the rate of false alarms.

One third discussed below Step in the thermal diagnostic procedure for axis-related elements refers focus on the detection of dangerous Conditions based on one or more BWBTIS measurements and others applicable information and data.

The condition that must be met for triggering an alarm or sending a warning regarding the measured temperature of a particular element may be written in generic form as follows Π (T HTDS , μ 1 , μ 2 , ..., μ N ) - Ξ (ν 1 , ν 2 , ..., ν M ) - γ ≥ 0 (141) where T _{HTDS is} the representative temperature assigned by the BWBTIS measurements to an HTDS for a given mechanical element, Π the function of T _HTDS and of N variables and / or parameters μ _n , and Ξ the function of M variables and / or parameters ν _m , where N and M are equal to or greater than zero. The amount γ is an optional term for the "Alarm level setting". Applicant states that condition 141 may be expressed in other formulas and that the above formula has been chosen to support the discussion below regarding hazard detection data processing methods. In particular, all dependencies of variables and / or parameters μ _n and ν _m and the value γ in a single function H could be summarized by rephrasing 141 as the following condition T HTDS ≥ H (μ 1 , μ 2 , ..., μ N ; ν 1 , ν 2 , ..., ν M ; γ) (142)

The simplest degenerate form of 141 equals Π as a multiplication of T _HTDS by 1 and Ξ as a constant value (which is tantamount to testing whether T _HTDS equals or _exceeds a be is a correct alarm threshold for the temperature). However, such a degenerate form is considered poorly applicable in the hazard detection problem discussed herein and some important dependencies are given below for the functions Π and, which serve to promote hazard detection sensitivity while keeping the rate of false alarms relatively low.

The function Π may embody one or more dependencies to transform the value of T _HTDS into a more accurate value, with one or more aspects involved in the process or processes (hardware and software) for determining T _HTDS from the fundamental measurements of the electrical BWBTIS signals.

Provided no dual wavelength (ratio) or Multi-wavelength thermometry used (when using special BWBTIS instruments), an important example of a correction of the function Π refers to the HTDS emissivity. The correction of emissivity requires the definition of a function by the spectral Sensitivity of the actual Meter and the nature of the observed surface. The Correction function is usually not linear. In our case can the HTDS emissivity for the different elements are very different (from high-emissive rusty elements too low-emissive shiny surfaces of chromed or polished metals and alloys) and may be of Age and history of the observed element depend (oxidation, peel the color, pollution, etc.). It is recommended that the correction in terms of emissivity to do so that one has the specific function parameters in the vehicle database stores and retrieves (e.g., using a polynomial Formula), namely for a special HTDS. Such a special definition of the correction can limited to those elements be, for the correction is more important. The actual correction parameters can be determined on the basis of specific statistics, or by defining a set of "reference corrections" for element classes (e.g., aluminum alloys, stainless steel, painted Steel from which the paint peels off and the rust can, etc.). The vehicle database is the information source thanks to which one decides which correction class for the relevant one HTDS of a particular element is to apply. Standard deviations can also for the correction parameter of the emissivity of specific elements or element classes (or for the corrected temperature), which, if appropriate, in the Π function considered become. A second example of a correction of the Π function is the compensation of the ambient temperature radiation at a HTDS with a low emissivity (and thus high reflectivity) and a Relatively high ambient temperature (the correction can either are based on the measurements of the ambient temperature or on the Readings of certain BWBTIS, if these heat radiation in the background measure up).

One third example is the compensation of the effect of the ambient temperature on reading the BWBTIS instrument when compensating for the temperature change of parts of the BWBTIS instrument is insufficient. This correction can by measurements made by temperature sensors, which are installed inside the appropriate instrument, and / or by measuring the ambient temperature by external calibration means (at a fixed and controlled temperature and position).

One fourth example is the compensation of the measurement offset by Use of external calibrators (for a firm and controlled Temperature and position).

The Function Ξ can is considered as a specific threshold of defect detection with which the value Π is compared will, if possible within a tolerance range expressed by γ.

It is well known to those skilled in the art that for the mechanical elements of most interest, eg axle bearings, under certain operative conditions (eg loading a wagon) and observing a certain operational history (ie driving in the last 10 or more Minutes at a certain speed), the variation of the temperature threshold for a given failure or incipient failure condition is dramatically reduced if the ambient temperature is considered to be the thermo-zero. In other words, the excessively high temperature of a bearing shell relative to the ambient temperature caused by (normal or dangerous) friction depends little on the ambient temperature itself. The main physical reason is that the dominant heat removal mechanism (convection) depends almost linearly on the temperature difference between the heated part and the ambient air. In addition, the variation of heat generation by friction does not depend strongly on the temperature. Thus, a simple definition of the function Ξ may be the sum of the ambient temperature up to a certain threshold, which may be retrieved from the vehicle database. If the system knows the axle load, an additional member, namely the function of the vehicle load, may be added to the function Ξ defined above. Such a value can in particular especially, but not exclusively, be proportional to the vehicle load by a multiplicative coefficient, and be proportional to or a function of the bearing shell temperature with respect to the ambient temperature, with the corresponding functional parameters stored in the vehicle database.

One another member could be defined when the system (from the train or from the information technology system the railway company) the information to the recent past Ride history could retrieve (Distance Distance and speed profile).

One special member could be inserted into the function Ξ, around the extra To consider the heating of the HTDS with a bearing shell, if the appropriate Wheel is very hot after a braking operation. This link would be from the actual Depend on the design of the axle unit and could therefore specifically stored and stored in the vehicle database.

professionals is also known to be an effective way of defining temperature thresholds for alarms for axis-related Components calculate this as the average of Temperatures of other identical elements on the same vehicle or even throughout the train, plus a permissible deviation that is considered multiple or standard deviation of these temperatures can be defined. The reason for that is that all these identical elements share the same story in terms of the traveled Range, the speed over time, the ambient temperature and for Brakes and wheels) the braking activity exhibit. When the elements on the vehicle side are open (e.g., in the case of a bearing cup), the averaging can still be more useful if you apply them to both sides of the train, so that the conditions such as solar radiation and wind are taken into account become. The temperature of elements that have a certain temperature in relation to the environment, or the one multiple or standard deviation from the average temperature Compared to other identical elements, must be of the Averaging be excluded. The identification of Vehicles in the method allows it, this averaging more effectively than the previous state technology, because the system has identical elements for the Select the mean value calculation safely can. Furthermore the system can set the criteria for the triggering refining alarms through the use of specific parameters, which are to be applied to a certain type of mounted axle and retrieving the system from the vehicle database.

in der Systemsoftware als statistische Alarmbedingung beim Überhitzen von Lagern, Bremsen und Rädern verwendet weren. Der Koeffizient K' in 143 kann aus der Fahrzeugdatenbank abgerufen werden (sofern er speziell bewertet wurde) oder er kann für Komponentenklassen definiert werden. Das Ausschließen von "abnormalen Elementen" aus den N Gliedern der Summierungen 143 kann durch einen anderen alternativen Algorithmus erfolgen, d.h. indem man die Mitglieder der Elementgruppen mit den höchsten Temperaturwerten T_HTDS ausschließt, bis die Bedingung 143 nicht für alle verbleibenden Elemente N gilt.Thus, given T _{HTDS, m} as the temperature _value T _HTDS for a member m of a group of identical elements, the following expression

used in the system software as a statistical alarm condition when overheating bearings, brakes, and wheels. The coefficient K 'in 143 may be retrieved from the vehicle database (if specifically evaluated) or it may be defined for component classes. The exclusion of "abnormal elements" from the N terms of the summations 143 can be done by another alternative algorithm, ie by excluding the members of the element groups with the highest temperature _values T _HTDS until the condition 143 does not hold for all remaining elements N.

in the Generally can other and more advanced statistical tests as a substitute for the simple statistical Approach based on variations can be used. Besides, different statistical Test for the same special item can be used, depending on the number of Patterns used in the analysis.

In the particular case of journal bearings, condition 143 may take into account some factors (eg, on-axle wagon load and radiant heating) that affect the temperature of the individual elements or subsets of elements by specifying T _{HTDS, j} , T _{HTDS, k} and T _{HTDS, i} be replaced by T * / B, h, as defined in the following expression T * Bra = T Bra - K''M H (T Bra - T amb ) - K '''(T. W, h - T Bra ) K '''' (144) in the T _{B, h} to the value T _{HTDS, h} for bearing h corresponds to T _{W, h} to the value T _{HTDS, h} for hours with the bearing associated wheel body is T _amb of the ambient temperature value, M _{h is} the total load or the net load on the axle or wheel associated with the bearing is h, while K '', K '''andK''''are constant parameters (preferably specific and retrieved from the vehicle database).

Applicant emphasizes that expression 144 is only a particular case of general mathematical expression T * Bra = T * Bra (T Bra , M H , T W, h , T amb , α 1 , α 2 , ..., α K ) (145) where a set of K parameters α _{n is} specific to a particular mounted axle or bogie model or vehicle model or classes thereof.

Some Types of BWBTIS can also provide measurements of the axle temperatures, and these measured Values can Be the object of a special alert criterion or in the expressions for alarm conditions warehouse and brakes are used.

The Member γ in The above expressions may be neglected or it may Fine-tuning the alarm condition can be used. This role of γ can however, by fine-tuning another, not necessarily additive, parameters within the alarm conditions discussed here or even in other conditions that are defined for the same purpose can be be replaced.

in the In general, more than one criterion for hazard identification for one certain type of axis-related elements are used, and the vehicle database can specify which criteria for a given Vehicle or swivel type or mounted axle type can, along with the parameters for the alarm expressions or the indication of the appropriate class of Axes, for these parameters have been defined.

in the Case of a complete high performance commercial machine for performing the Detection of axis-related thermal defects, the system can the processing unit of this device a set of vehicle-specific Provide information that is useful could be, to make its own diagnostic criteria more specific and effective To apply. Such a procedure would be a special form of Implementation of the above.

Of the Value of the ambient temperature used in the calculations for the diagnosis the o.g. Axial hazards can be used by various standard means are achieved, and in particular by a reasonable temperature test or even a local meteorological Station whose other data for other purposes useful could be.

In The past has been much concerned with the problem, like one can distinguish sleeve bearings from ball bearings to different ones Apply alarm criteria or define alarm criteria that for both device types would be applicable. In this context, the applicant clarifies that such Aspect in this text especially therefore not discussed in detail is because socket bearings are practically no longer used (see above) are they e.g. in Europe under the RIV agreement in vehicles not more allowed). However, the method and the system would be the question of the two Solve bearing types in a brilliant way, since the identification of vehicle models the possible availability any useful information for it implies, including Information about the bearing type (and even the model, the geometry, the position etc.), as explained above has been.

5.12 Overheating diagnosis and fire in boxes of identified vehicles

Of the following text deals with the diagnostic functions that are in the system can be implemented to an overheating or to detect a fire aboard identified vehicles, based on the passive thermal infrared measurements, the Calculation of vehicle position over time and specific Information that can be stored in the vehicle database.

5.12.1 Sensors for diagnostics from overheating and fire from boxes of identified vehicles

To detect overheating and fire aboard identified vehicles, the method requires a number of appropriate measurements of the thermal infrared radiation emitted by the car box or its load to process them as follows. The requirements for these measurements are very similar to those discussed above in the thermal diagnosis of axle-related components, although in the case of vehicle bodies and loads, the requirements for spatial resolution (based on the positions of adjacent measurement points on the observed surfaces), thermometric resolution and accuracy, the measurement bandwidth and pattern rate may be less binding. In this case as well, the applicant does not give strict specifications for the corresponding passive infrared sensors, but it will be clear from the explanation of the data processing method below how the characteristics and installation parameters of such sensors improve system performance in the detection of overheating and fire aboard affect identified vehicles.

21a and 21b show a possible positioning of a linear passive infrared imaging device 760 or 768 , which performs its measurements for a vehicle side and the upwardly facing part of the surface. 21b is a slightly different view of the installation of 21a since the positions and orientations of the measuring instruments shown are not the same (angle 772 is zero 21a ). In the two pictures 21a and 21b only one infrared imaging device is shown, but of course another corresponding imaging device would also be installed symmetrically for the other vehicle side. In addition, other imaging devices could be installed in different locations and orientations.

When linear infrared imaging devices are used, the gap hangs 767 between two measurement points corresponding to the adjacent pixels (usually from one of the two corresponding measurement beams 763 and 764 , obviously from the distance between the observed area and the imaging device. For example, if the above-described thermoelectric linear "Model IR 1000" [068] with a field of view (angle 765 between the extreme measuring beams 761 and 762 ) of about 65 deg, the approximate length value is sufficient 767 for the middle pixels of the die from about 13 mm at 1 m distance to about 65 mm at 5 m distance. These two length values 767 for a 256 pixel linear imaging device, about 5 mm and 25 mm are the same under the same conditions. In contrast, the longitudinal distance between two successive imaging beams depends, as well as the distance 775 between the rays 773 and 774 , only based on the sampling rate and the vehicle speed (for example, the approximate value is 33 mm at 120 km / h with a sampling rate of 10 ³ s ^-1 ). Of course, the actual skew angle between the measuring beams and an observed surface defines the corresponding distance between the measuring point centers on the surface. As explained above for other instruments, the reduction of the angle affects 772 from 90 degrees to the minimum value of the angle 765 for a complete view of a range of positions on the vehicle. Of course, the field depth of the optics plays a role, and the optical resolution of two adjoining pixels should be suitable for the corresponding distance, and also the longitudinal distance 775 must be taken into account.

Passive IR scanners based on rotating mirrors (eg, the modified version of the fast scanners [964]) are an alternative to linear imaging devices based on a rigid die and may be considered for system implementation. As discussed above for the other instruments (eg the VLDS), the choice of the number of units and their positions and orientations for the system implementation is a result of balancing the system performance in relation to the expected costs and maintenance, with the characteristics of the instruments ( Resolution, sampling rate, etc.) and the installation constraints (especially in the case of installing double or multiple bus bars). Applicant believes that the installation of two linear imaging devices at a similar position as in 21a is sufficient to obtain valuable results by the data processing method described below, and above all, both of the above-described linear infrared imaging devices based on a thermoelectric linear matrix or photoconductive PbSe matrix or fast infrared scanners have been successfully used can be used to carry out the thermal infrared measurements of vehicle boxes and their load. However, it is possible to use other types of linear infrared imaging devices, or FPA infrared imaging devices, or even series of instruments with one or more thermal infrared sensors for each unit, provided that their measurement bandwidth, measurement rate, field of view and resolution are adequate and Each measurement can be assigned an exact time value. 21a and 21b also show the VLDS instrument 766 or 769 as well as a linear camera (VIS or NIR or color camera) 770 , The installation of the three sensor devices (linear IR imaging device, VLDS and linear camera) or two of them (linear IR imaging device and VLDS or linear IR imaging device and linear camera) with parallel view planes and with measurement beams on one same line converging parallel to the rails is particularly recommended, as explained further below, when the data they generate is the object of a common process. In fact, with such an instal the corresponding data for these different instruments are easily matched (based on the Ω coordinate transformation, or only on LDF, together with the calibration parameters of the Instrumenengeometrie) and offer the important advantage that the relative positions of bodies in the foreground and background of the vehicle from the same relative point of view (at different times).

5.12.2 Data processing for the Detection of fire / overheating on boxes identified vehicles

The Processing of data of thermal radiation measurements for boxes of identified vehicles (including their load) to overheat and to detect a fire hazard based on the information which are contained in the vehicle database, for the respective Vehicle model actually defined diagnostic method and the vehicle-specific Data for the application of the methods provides. Such methods are here labeled "VBTHDM" (for Vehicle Body Thermal Hazards Diagnostic Method). Each VBTHDM uses one or more "TEPP" (used here Acronym stands for Thermal emission data pre-processing algorithm), some numerical Deliver values by taking data from a particular subgroup of thermal radiation measurements which correspond to a "TESD" (the Acronym stands for "Thermal Emission Spatial Domgin ").

In general, a TESD is defined by an area in the C _VB coordinate system and corresponds, more or less accurately, to a physical area that is a fixed part of the vehicle body or a physical area of an element on the vehicle, without a given fixed position, or a specific "virtual envelope" defined for the vehicle. A TESD1 is a geometric surface corresponding to a physical surface associated as a permanent physical element with a particular vehicle model. TESD1 surfaces are defined in the vehicle database as rectangles or more complex structures such as a mosaic of simple polygons such as triangles, a flat surface with a polygonal outline, or an analytical geometric surface such as parts of cylinders or cones. Some examples of matching a TESD1 with a physical vehicle element are a flat sidewall of a closed boxcar, a closed sliding door for the same type of coach, a window of a passenger car of a train, the flat top of a closed car for transporting grain in the form of Bulk goods, a cylindrical surface of the roof of a closed freight car or a blind wall on the side of a locomotive.

A TESD2 corresponds to a physical area that is specific to one Vehicle was detected by the system (due to measurements) and that does not match an element, this is a permanent component with a fixed position in the corresponding one Vehicle model is. TESD2 surfaces are usually called mosaics or as a series of three-dimensional Outlines or profiles defined. Some important example of TESD2 are tarp shrouds or covers of loads on open wagons, the surface of heavy bulk cargo on open waggongs, solid bodies mounted on flat rail vehicles are loaded, and vehicles transported on flat wagons or cars on double-decker wagons. A TESD3 surface will used to define a geometric envelope that no one corresponds to physical element whose processed heat radiation data but useful are. A TESD3 is defined by surface elements, as defined by TESD1. A notable example of a TESD3 is a part of a loading profile serving for one open wagon. One or more TESD3 surfaces together can be one Define TESD3 volume that is functionally equivalent to the corresponding one Group of TESD3 surfaces is.

The Thermal radiation measurement data become over with a TESD the Ω coordinates defined above, or defines the LDS as a degenerate and limited case the Ω function can be seen. Every heat measurement corresponds to a measuring beam, which (in whole or in part) one or several TESD areas can cross. The calculation of the interfaces requires the Application of said coordinate transformation function, the geometric one and calibration parameters of the corresponding thermal radiation measuring instrument as well as the TESD data from the vehicle database. The uncertainty when determining such interfaces should be considered as well as above for the processing of the BWBTIS data was explained.

A measuring beam can be aligned with one or several TESDs, depending on their full or partial interface with such surfaces and on the "transparency" of the corresponding TESD surfaces. The "TESD Trasparency" can be defined as a number between 0 and 1 that expresses the average fraction of infrared radiation that is likely to pass such an area. The corresponding "languor" of a TESD is the complement of the transparency of the entity. It can be a minimum and a maximum Transparency value for each TESD. The transparency can also be defined as a function of the beam tilt direction in the TESD. TESD3 surfaces are completely transparent, while the TESD1 surfaces are basically completely dull, unless a transparency has been specified (eg in the case of a lamellar plate or a metal grid). The transparency of TESD2 surfaces could, if necessary, be determined by an algorithm defined taking into account the properties of the corresponding three-dimensional measuring system and the thermal measuring system (with special reference to the space between adjacent measuring beams and the interface of the beams) two types of instruments). For each measurement beam, a set of "assignment coefficients" (also called "BAC" for Beam Assignment Coefficient) can be calculated by determining the intersection of the beam with a series of TESD surfaces onto which this beam originates from the measuring instrument and towards the background A "fraction of the interface" (also called "BIF" for Beam Intersection Fraction) can be defined (with more or less complex algorithms) to quantify the relative interface of a measurement beam with an area. the unit value being large if the interface is perfect and proportionally smaller if it is a subinterface; When doing so, the actual beam profile should be taken into account (as well as the relative uncertainty of the beam direction and its positioning.) A beam relative integrity equal to 1 is assigned to the beam before each TESD surface interface The BAC of a TESD surface is determined by BIF multiplied by the radiographic integrity that the surface achieves, if the transparency is represented by a range (and not by an area) single value), different criteria may be used to assign one or more BAC values, BAC values are generally considered to be "fuzzy" variables, except for the unit value, which is very common and a full assignment of a TESD variable. Area corresponds to a measuring beam.

Of the Applicant clearly states that this formalism, which uses BRC values, about matches of measuring beams and surfaces To determine a practical way is to ask the question Can formalize interface of the measuring beams with TESD, however he by other formalisms with an equal, lower or higher complexity level be replaced. The process of measuring a measuring beam of a several TESD areas can be limited to it be, whole or partial matches (with or without the definition of BRC values) or he can also include the calculation of the coordinates of the interface center. The interface profile on the surface and corresponding variance values may also be be calculated, if you are for a TEPP are required.

The temperature values obtained by infrared measurement of heat rays can be used directly in the TEPP calculation, or, in the case of TESD1 surfaces, they can be corrected as discussed for the T _HTDS measurements of BWBTIS instruments also to consider the emissivity and other factors (based on the corresponding information associated with the TESD1 areas in the vehicle database). TEPP1 is a simple way of processing thermal radiation measurements of a matte TESD1 surface and consists of calculating the average value (and possibly the standard deviation or other statistical moments) for the measuring beam with BAC equal to 1, ie when the beams are unrelated non-transparent surface coincide. TEPP1 may thus be ideal for TESD1 surfaces of sufficient size (in terms of the distance and beam profile of the infrared heat measurements) for which a fundamental temperature homogeneity is expected. TEPP1 thus corresponds to the TAM1 method explained above for the HTDS of axis-related elements.

TEPP2 is applicable for relatively small TESD1 areas, if no relevant measuring beam has a BAC value of 1 The above defined methods for TAM2 can be applied in this case to a representative temperature value calculate and, if possible, for one corresponding uncertainty estimate.

TEPP3 corresponds to a partially transparent TESD1 surface in front of another TESD1 surface and is applicable to measurement beams with BIF equal to 1 for both such surfaces. The statistical distribution of the measured temperatures is calculated and the average is determined from two sets of measures containing the lowest N _L values and the highest N _H values. N _L and N _H are specified for respective VBTHDM or calculated by an algorithm taking into account the entire pattern size. TEPP3 is applicable to "semitransparent" surfaces with slits or holes and fixed parts whose minimum width is greater than the maximum measurement point diameter in the front surface measurement of the radiation (therefore it is assumed that some of the beams are only one of the two TESD1s). Cut surfaces completely or partially).

TEPP4 is applicable to the same case as TEPP3, but it is based on predicting the average temperatures of foreground and background, assuming a certain beam profile on the foreground surface TESD1 and a particular structure of the same area (typically referred to as a panel with alternating slots, stripes or holes is described becomes). TEPP4 is applicable to "semi-transparent" surfaces with Slits or holes and fixed parts whose minimum width is less than the maximum elliptical diameter of the beam spot of the heat measurement of a surface in the foreground.

TEPP5 is applicable at relatively large TESD1 or TESD2 matte and flat surfaces on which one or more warmer Areas can occur while some other areas have a homogeneous lower temperature. Cylindrical surfaces can in equivalent flat areas being transformed. The measuring point centers are first in two dimensions on a flat or flattened TESD surface recorded. So a pretty good card or a "pseudo-image" is defined and compact accumulations of warmer Points and cool Points are identified and determined by suitable algorithms. A possible Algorithm is based on the definition of a set of binary representation limits (e.g., based on a global histogram of temperatures) followed by the generation of binary mappings to which the blob analysis is applied. Other alternative algorithms can be used as well as approaching through "Bi-Gauss" functions (temperature across from the orthogonal coordinates of the TESD surface, where the axis of the elliptical outlines is rotated at a certain angle, to be adjusted) after searching for local maximum and minimum values (around the coordinates of the bi-Gaussian function centers to define) on a map by a two-dimensional Adjustment of the thermal data card was generated. TEPP5 requires the definition of some parameters to the search for warm and cool Restrict posts (e.g., minimum accumulation size, maximum corresponding temperature range in absolute values or in relation to the Ambient temperature or range of temperature values, etc.). The output of TEPP5 consists of a number of representative ones Temperatures (maximum and / or average) of hot and cool spots, preferably along with their coordinates, the elliptical diameters and one or more measurements of statistical dispersion.

TEPP6 is equivalent to TEPP5, but limited to higher temperature ranges of corresponding TESD.

TEPP7 is equivalent to TEPP6, but it is applicable to TESD2 surfaces and that took into account Temperature readings are those that exceed a certain percentage of the corresponding histogram.

TEPP8 is defined for the same data recording as TEPP5, but it will be a histogram of Temperature values created after a two-dimensional adjustment. The histogram is preferably normalized by the surface areas, based on a single measurement based on the average distance refer to the adjacent measuring point centers.

TEPP9 is for TESD1 matte surfaces defines and serves the presence from warmer Detect temperatures on one or more lines on the surface. TEPP9 can be based on different formulas and calculation processes (Calculation of co-variance, pattern recognition, etc.), and the lines can be in the vehicle database are determined by their position, or even by looking at their minimum length and relative orientation limits. VBTHDM1 is a relatively simple method, based on the average Temperature values of TEPP1, or at the point temperatures of TEPP2 or TEPP6 or TEPP7 or TEPP9 or at the foreground and background temperatures TEPP3 or TEPP4. Alarms are triggered when a high temperature during processing of the corresponding TEPP (at the given limitations and if the processing parameters are defined for such a TEPP) the ambient temperature exceeds a certain threshold, the for the corresponding TESD areas a vehicle model is defined. VBTHDM1 has the advantage that she for a single TESD is applicable without requiring a different TESD for reference available have to be. However, it is also limited by the fact that the temperature variability due to sunlight or due to heat sources in the vehicle considered can be, resulting in lower sensitivity within the admitted low rate of false alarms.

VBTHDM2 may be based on average temperature values of TEPP1, higher or lower temperatures of TEPP2 or TEPP5 or TEPP6 or TEPP7 or TEPP9, or on the foreground and background temperatures of TEPP3 or TEPP4. Alarms are triggered when at least one high temperature has been removed from processing by the corresponding TEPP (with the limitations and limitations) Processing parameters defined for such a TEPP) have a reference temperature value calculated by the same TEPP or other TEPP and applied to the same or different TESD areas by a certain threshold. One can define the low reference temperature by taking the average of the values of one or more TESDs, excluding as far as possible the values that lie outside a certain variance multiplier. VBTHDM2 may make it possible to improve sensitivity compared to VBTHDM1 with respect to the effect of solar radiation when the reference and alarm TESD have the same orientation, but such improvement will be due to the differences in heat conduction and emissivity of the corresponding ones Restricted areas.

The applicable differential thresholds for VBHTDM1 and VBHTDM2 may be special ones be given values, or they can by a function of Standard deviation of the corresponding TEPP and possibly by a function the ambient temperature or a lower temperature of one TEPP be determined.

VBTHDM3 is based on the processing of the histogram resulting from TEPP8. Various alarm criteria can defined as a fixed or calculable threshold for the Difference between the average temperatures of certain solid or calculable percentage intervals.

VBTHDM4 based on the statistical meaning of the difference between higher Temperatures derived from a TEPP (especially from TEPP1 or TEPP2 or TEPP3 or TEPP4 or TEPP5 or TEPP6 or TEPP7 or TEPP9), and the average of the corresponding values for different TESD one or the same vehicle or a set of vehicles of the same execution model or part of a particular class of execution models. The statistical Meaning can be based on a standard statistical meaning criterion To be defined. You can do a similar one Apply a condition such as Expression 143. VBTHDM5 is based on the application of rules and mathematical expressions that the Alarm conditions based on more than one TEPP result define. For example, one of the two thresholds for a Couple of lower and higher Temperature outputs of the same TEPP or of two different ones TEPP uses, depending on the temperature output range of a TEPP for one certain TESD.

The Applying a TEPP to a TESD2 surface requires an appropriate Procedure is applied to the three-dimensional data that the area normally define an unknown load on an open one Wagon corresponds to process and that suitable geometric Restrictions by The parameters of the relevant VBTHDM are set to the TESD2 within that defined by said three-dimensional measurements total area to determine. TESD3 surfaces can used when the instruments installed on the SMI for the three-dimensional Measurements are not suitable to the area of a load on an open To define wagon. In this case you can VBTHDM1, VBTHDM2 and Apply VBTHDM5, provided that the high temperature values considering the for a TESD3 in the foreground and a TESD2 or TESD3 in the background given limitations have to be searched. In this way, the measurements relate either to the load (or a waggon sheet) or on the background of a wagon surface or a track area equivalent. The pure criterion of a comparison of the absolute temperature output from a TEPP (e.g., an average temperature of TEPP1 for a given TESD) with a certain absolute temperature threshold could be the basis for the definition of a VBTHDM, but the applicant holds this criterion for hardly applicable when compared to the VBTHDM alternatives described below compares. In fact, would in most cases the use of an absolute temperature threshold gives a low detection performance mean that you avoid too high a rate of false alarms would.

Of the Applicant makes it clear that the above definitions are more alternative Types of TESD, TEPP and VBTHDM are not restrictive and are not exhaustive. Other TESD, TEPP and VBTHDM can (by experts in the field addressed) for the implementation of the system provided that they comply with the principles in force of the procedure.

following be some considerations on the use of certain combinations of VBTHDM, TEPP and TESD for the main types of rail vehicles hired, with the Aim to illustrate and determine the application of the invention advices for the application certain areas of the procedure.

Fires erupting inside closed wagons with various freights and eventually reaching a potentially very dangerous level, may evolve over time with a very variable rate of change and localization of heat release, which depends on the source of ignition, the contents of the wagon, the gaps between the charged elements or between goods and their containers, the three-dimensional Charging pattern and response to fire, the thermal conductivity of the wagon walls and the ceiling and the ventilation of the hold. It is especially possible that for several 10-minute intervals after ignition and for a few minutes after the possible generation of a possible spark sparks, a fire will not cause the collapse or burn-out or very high temperature of any large part of the walls and ceiling within the loading compartment triggers. In this case, VBTHDM1 can be used with different TEPPs for specific TESD1 surfaces, such as vents, or with TEPP6 for whole walls, panels and doors, the main limitation being that the alarm threshold is set to: the strongest possible solar heating does not trigger an alarm.

VBTHDM2 Using TEPP5 is a very effective solution, which is quite a simple one and fast definition of the relevant TESD areas implied. Especially TEPP9 can be very useful for wagons with insulating wall panels with metal frame (which is a thermal bridge to form the internal temperatures) to be seen from the outside are. Semi-trailers and containers are usually similar in terms of their fire dynamics and suitability for certain detection methods closed wagons.

closed Wagons for transport of combustible solids in the form of bulk material (e.g. wood chips etc.) develop flames with relatively low heat release and with a higher probability Smoldering fires that are very slow and over a long period of time can develop. A differential temperature criterion like that of VBTHDM2 or statistical Criteria like from VBTHDM4 can prove to be effective to increase the temperature on the side walls locate while VBTHDM1 with TEPP1 or TEPP6 for the upper surfaces or the side walls can be applied.

Refrigerated cars represent a special case because of the possibility that the cooling unit the fire source is, and because heat exchangers usually a higher one Temperature have as "passive Elements "on one Wagon. Therefore, the localization of heat exchangers should be considered be pulled and some TESD can be defined to be a specific diagnosis of fire that is connected to a Compressor unit arises to formulate. Some refrigerated cars use polyurethane as insulation material and are special dangerous during a fire in a tunnel, as significant amounts of hydrocyanic acid through the Polimer pyrolysis can be released. The widespread preference from low-emissive Exterior surfaces, however, allows the increase the sensitivity of the detection method.

Car transport wagons and HGV transport wagons are characterized by the variability of the load and a relatively high fire hazard. An optimal sensitivity setting for this Wagons hangs from the distance of the SMI installation from the nearest loading point because of the possible Residual heat in Tire, engine, exhaust pipe and brakes, if that transported Vehicle was stopped just before the system installation happened.

Fire On passenger rail vehicles are principally by arson caused and they can too dangerous Develop situations, depending on the fire behavior of the building materials (especially the seat cushion) and the luggage. Some of the above for the closed Wagons made general comments also apply to this Rail vehicles, but with some additional information. window frame usually behave like thermal bridges and it may offer apply TEPP9 here. Especially simple window panes up faster than sidewalls and can for the Definition of TESD1 surfaces be used because their low transparency in the range of thermal Infrared wavelengths (emissivity is close to 0.8 for most slices and their variance is very low) recognizing abnormal overheating allow. The presence of heating systems should be considered be defined by defining certain parts of the car body Excludes TESD surfaces, so a higher fire detection sensitivity is reached. The fact that at low outside temperatures the heating is on, limits the sensitivity of VBTHDM1. The use of air conditioning can lead to, that a certain part of the car body is colder, and this should be at the definition of TESD and TEPP with regard to the lower temperature reference considered become. In sleeping cars, the risk of fires is not through Arson caused larger, since a relatively high density made of flammable materials as they are divided into compartments and because smoking in bed can cause sparks to jump over. For her Nevertheless, the same consideration applies the detection methods as for normal passenger cars.

locomotives represent a very special case and are characterized by a relatively high fire hazard, however, from their model and general depends on the type and generation of traction. Put diesel locomotives a particular danger in tunnels because of the combination of fuel and danger of ignition. The danger of ignition does not exist mainly in the context of the fuel tank, but with the engine compartment, especially because of the possibility that diesel fuel escapes under very high pressure because of a leak. Diesel locomotives also have a number of hot and hot areas the engine and exhaust elements, and even if the engine surface of a Diesel locomotive enclosed in a housing is, so it is partially visible through the ventilation panels and can be the cause of a false alarm. Electric locomotives generally have no drive components that have high temperatures but it can be reduced by the removal of heat, e.g. from the electrical Brake system, too high temperatures on its outer surface come (especially after a long and sloping rail route). Because of these features should be the definition of VBTHTD, TEPP, TESD and the corresponding Parameters for every locomotive model will follow exact and special considerations about the technique, and a vote of the recognition process is highly recommended. The More complex TEPP calculations and the use of VBTHDM5 may be possible be required to have very high detection sensitivity for fire and overheating while maintaining a low false alarm rate.

All Types of vehicles that are overheated by the methods described above. and fire diagnostics require the definition from VBTHTD, TEPP, TESD and the corresponding parameters and that is the the procedure most likely to be recommended is determining a first such Definition, which will then be corrected one or more times. The availability of thermal cards from the system can in this first stage the definition of the method for a particular model will be very valuable and it is useful the cards for to save all false or real alarms to the success of the Increase recognition process. The fine-tuning of detection methods through optimization the TEPP parameter can best be done Becoming a big one Number of measurement records stores and the method has been revised offline, while one further tuning the parameters such as differential thresholds appropriately, when storing and analyzing the TEPP outputs.

Of the Applicant also notes even if the topic is not discussed in detail in this text, Professionals can develop software algorithms and applications, thanks of which the fine-tuning of the methods is "almost automatic".

If some individual vehicles, e.g. one or more locomotive elements within a row, trigger an alarm while other vehicles of the same execution model If you do not do this, you can change the applicable hazard detection method and the changed Apply method if the serial number thanks to the unmistakable Vehicle marking is detected, which for most of the vehicles by the Reading process of the marker codes in the vehicle identification procedure described above is determined.

5.13 Boundary line profiles and thermal diagnostic methods at unidentified vehicles

The Diagnostic methods described above can be high performance or Sensitivity within the required low rate wrong Alarms reach you thanks to model-specific information and data and thanks to the accuracy of associating measurements with certain known parts of a vehicle execution model. The text below discusses some options that are in the system in terms of diagnostics capabilities for small parts of vehicles can be implemented whose model is not identified by the appropriate procedures has been. If not the decision is made on the application certain diagnostic functions or all diagnostic functions others can do without unidentified vehicles Methods are developed and applied, even when compared to the methods for the vehicles whose model has been identified are less effective are.

The assessment of boundary line profiles from the box of unidentified vehicles can be performed using a modified version of the methods described in Section 5.9, based on a coordinate transformation function Ω, less accurately and safely calculated, or by a modified version of the method described in Section 5.8. The simplest way to calculate the parameters of a less accurate Ω function is to limit the longitudinal displacement defined by the LDF function. Defining the other (angle and displacement) components may be done based on the detection of a particular vehicle body element at different times along the SMI. Detecting the same profile on both sides at certain appropriate heights can be particularly useful be to determine the lateral shift, the gravity value the vibration value. The major components of buffering are a particular example of an element that can be used for such a purpose. Accurate measurement of the lateral displacement of the wheels at a plurality of longitudinal positions along SMI may be useful in calculating the lateral displacement over time of bogie boxes or centers of independent (non-bogie mounted) axles. The permissible profile of the vehicle body (under load) for use in one or more hazard detection conditions, such as 126, 128, 127 and 129, may be calculated according to the methods of UIC-505-1 standard [050] or similar methods. The most important input for such a profile calculation corresponds to the positions of the bogies or the centers of individual bogies (not mounted on the bogie) known from the LDF and WSD calculations. Other input data, such as the Vehicle Flexibility Coefficient, may be suspected, possibly with the worst-case assumptions.

The Diagnosis of boundary line profiles of the lower parts of a Vehicle can for Unidentified vehicles with only a few disadvantages in comparison a method performed which is based on vehicle recognition, as the wavering and the aspects of non-standard load are less important.

The Diagnosis of overheating, Failure and incipient failure in axle-related components of Unidentified vehicles can have no advantages over the provide model-specific information and data.

However, one can develop data processing algorithms (largely dependent on the BWBTIS installed) that perform the diagnostics in a more or less sophisticated manner. The Γ _WS function can be determined with great accuracy at least for the longitudinal displacement and yaw components from the wheel sensor data and then used in the BWBTIS processing. The assignment of BWBTIS to certain elements such as bearings, wheels, brake discs and axles can use the processing of the VIS or NIR images. Measurements of fast and accurate laser distance meters may be useful to determine the wheel radius that can be defined for defining the height of the bearings above the rolling surface and the lateral displacement of the wheelsets. The diagnosis of fire and overheating, based on measurements of thermal radiation for the boxes of unidentified vehicles, may use the Ω function, which, although defined solely by the LDF function, is generally sufficiently accurate for this purpose, and especially if there is no model-specific information and data. Methods such as VBTHDM1 with TEPP1 or TEPP5 or TEPP6 can be used, even though the thresholds, TEPP parameters and TESD definition can not be translated into a specific model and hence a decreasing diagnostic sensitivity at the required low rate is more incorrect Alarms results.

in the Principle could at this point an undissolved List of vehicle model candidates from the vehicle identification application used when looking at the information and data for this Considered candidate models and selects them logically or the result obtained by their use logically with the help diagnostic methods that filters for identified vehicles were determined.

5.14 Specific functions, which concern the transport of dangerous goods

Some specific functions can in the procedure and in the system in relation to the transport of Dangerous goods are integrated, especially in connection with the Generation of an information record for each train and transmission this data and other (information) systems or saving in the System and retrieve the data as needed.

Some Dangerous goods are often mass-produced by train on a train transported by special tanker rail cars, which are dangerous with certain chemical substances, flammable liquids and compressed Gases are compatible.

Intermodal transport is also widespread with tanks on transport carriages on open railcars, with containers and with semitrailers on bogies, even if certain special goods are excluded. Hazardous goods are also transported in their own special packaging on normal closed freight rail cars or by special rail cars (for example in the case of radioactive material). In most cases, it is mandatory in most countries to attach special standard labels or markers or posters to the pages of the corresponding wagons. Thanks to a number of international agreements, these markings are the same in many countries and generally contain one or more readable mark codes specifying the dangerous content and the associated hazard or dangers. To this day, the United Nations has various agreements for the transport of dangerous goods worldwide, and the individual dangerous goods are often identified by their "UN number". In the case of rail transport in Europe and other Member States of the "OTIF", the "RID" agreement [061] regulates a large number of detailed technical issues.

The Detection of Danger Signs can be done by the system using of image processing functions used for the images of vehicle pages applied become. In particular, you can For this Purpose the above explained VIS or NIR linear cameras are used when looking at the image data arrangement as explained above produces and if possible the information from the vehicle database used to the area restrict in which the mark is to be recognized. Color line cameras or FPA cameras or B / W cameras with spectral filters can be used Be sure to get the special orange color that is available for most today the markers mentioned here is used to their advantage to use. The search for the dangerous goods marks can be identified on those Rail vehicles limited be that for such transports permitted are, and on all unidentified vehicles, thus giving an unnecessarily large computational burden avoids.

in the Case of Tanker Railway Cars are special hazardous goods posters in general both for filled Tanker as well Tanker required for used the transport of a certain dangerous goods and after were not cleaned during transport. It is therefore possible and useful, the appropriate date indicating the actual amount of dangerous goods to add in the tanker, by taking the weight of the empty vehicle (from the vehicle database available) deducted from the gross weight, which one for the appropriate vehicle can calculate when installed a load meter for the wheelsets and in the System is integrated.

Also if the information of presence Of dangerous goods on a train often already by various others Sources available is, and even if in the next Years increasingly in many geographic areas locating systems are implemented, so can the implementation of the above-mentioned functions for transport of dangerous goods useful be an easy to read and automatic source of others and to get redundant information specifically related with managing accidents can be valuable in tunnels, as recommended in some recent security policies (e.g., [063] and [064]).

The as above generated information about rail transported Dangerous goods also of railway traffic control systems for one redundant control can be used if driving in a tunnel a freight train that is any or a specific dangerous goods transported, not allowed, if at the same time There are other trains or passenger trains in the same tunnel.

5.15 Integration of measurements of wheelset loads and wheel error detection

Of the Previous state of the art knows various solutions for the detection of flats of wheels based on measurements using different types of Acceleration, force or deformation sensors [032, 036, 039, 040, 041, 067] mounted on rails or ultrasonic sensors [034] or by multiple optical detectors [038] or by the acoustic one Feel the periodic stops [035] or by electronic means for sensing contact loss between a wheel and the rail [033, 037]. Recognize some of these solutions also other wheel defects, such as a "welding". Some systems have also been developed and put into the trade to wheels, Wheelsets, bogies and to weigh whole railcars while the rail vehicles drive past a measuring point and some of them (e.g., [040, 966, 967]) combine the load measurement function with the wheel defect detection. Wheel Defect Detectors and Load Measurement Systems will be installed both for improving the safety of trains as well as reducing them the maintenance costs for Rails and rail vehicles. Detecting flats allows this Turning the wheelset or changing the wheel immediately This results in less rail wear and less Risk of accident, e.g. at a rail break, especially at low Temperatures. The detection of an excessive load on an axle and an imbalance of load between wheels in a wheelset already used to deteriorate the rail condition prevent and reduce the likelihood of accidents.

Wheel Defect Detectors and load measurement systems can be easily integrated into the system, in Install sensors at the SMI or close to the SMI and send the appropriate data from the wheel defect diagnostic data processing equipment and / or load measurements to the system's computing equipment. The investment for the development of such an integration is very small and the data transfer can be done in many different ways (BUS, LAN, etc.), even after the characteristics of the devices used for the wheel defect diagnosis and / or the load measurements. The data acquisition equipment for the sensors used for wheel defect detection and load measurements can also be integrated into the system's data acquisition equipment and can provide the data processing application (s) for wheel defect detection and weighing on one or more of the system's data processing units let run. This latest integration plan requires more development investment, but it has the advantage of reducing recurring costs. In general, the association of the output of the wheel failure detection and / or the wheel load measurement systems with the wheelsets that are independently recognized by the system may be based on a serial number of the wheelsets or by passing time, if the possible time difference between the wheel sets different systems is low enough, or if it is known with a sufficiently high accuracy.

Some Advantages of integrating the wheel load measurements into the system are related with the possibilities to use such measurements combined with other data that the system determines or calculates or from the vehicle database retrieves, and thus improve system performance. A first advantage is the use of the wheel load data within the vehicle identification procedure, which was described in section 0. A second possible Advantage results from the application of vehicle load information in the data processing application described in section 0 above was to detect dangers of boundary line profiles, and especially for the use of the actual Load instead of the maximum load in the calculation of the corresponding Values of Expressions 126 and 128. A third potential benefit is the use of wheel load data as discussed in section 0 above, to improve sensitivity in error detection of Ball bearings at a low false alarm rate. Other types of security alerts can generated by the system when the wheel load measurement is integrated, by the load per wheel set, the load per bogie, the load per wagon and the imbalance of the load with specific thresholds that are derived from the vehicle database for the Wagons whose model has been identified can be retrieved.

you may be certain data that the system collects or calculates or retrieves from the vehicle database (e.g., WSD and wheel diameter) as well use the performance of the software application in wheel defect detection to improve.

One Another important advantage of the integration of Raddefekt detection and the load measurements in the system is a total savings since the same devices and the installation for the data transfer, the signaling, the wiring, the power supply, the gear boxes and other additional Infrastructures are used.

5.16 Installation of the System sensors on a curved railroad track

professionals the appropriate technique is clear that in this document explained Methods easily changed in detail can be to install the system on a railroad track curved in the area of the SMI, and that respect the choice and positioning on the rail bend and the related inclination the rails are not big Problems occur.

What As far as the LDR and WSD calculations are concerned, it is recommended that the Definition of the LDF function (relative to the wheel center) over one curved Axle following the rail center between the rail tracks, define. An equally recommended choice is the one that covered by the individual wheels Distance, as well as according to the longitudinal coordinates of the individual Wheel sensors on such a central curved axis based on the corresponding transverse sections orthogonal to said curved axis lie to determine.

The introduction of the rail curvature into the function discussed above for the definition of the function Ω has some small practical consequences. In particular, one can use the same C _GB and C _VB orthogonal coordinate systems as above and can formulate a first estimate of the components of the Ω, and especially the values of ψ (t), Y (t), and Z (t) using the LDF and the curved longitudinal axis forming its area, as recommended above. The first estimate of θ (t) may correspond to the inclination of the rails, or may also include the average effect of vehicle speed and possibly the flexibility coefficient from the vehicle database (for identified vehicles) into account. Similar considerations can also be made for the definition of the function Γ _WS discussed above for the axis-related elements.

Any other change Details of individual methods and design features for the procedure and the system in connection with the system installation, though the SMI on a curved rail track should be based on the design of the implementation by experienced engineers the basis of the explanations for the Procedures in this document can be easily defined.

5.17 Integration of others Sensors or sub-systems for additional security features

Other subsystems, instruments, and methods may be added to those discussed above to provide further diagnostic capabilities and possibly combine the various diagnostic elements to synergistically enhance certain methods of detecting erroneous and hazardous conditions of passing train vehicles , Some diagnostic functions related to the current collector of electric locomotives can be integrated into the system, for example by applying the solution, which was developed by AEAT in Derby, United Kingdom, for their system "PANCHEX ^®" [968]. Additional diagnostic features for pantographs can be designed and implemented by using VIS or NIR cameras and / or IR matrices or scanners with appropriate resolution. The use of the LDF or Ω coordinate conversion function together with the data and information retrievable from the vehicle database for the corresponding locomotive makes it possible to perform automatic inspection of the pantograph geometry and possibly pantograph wear. The estimation of current collector temperatures (especially of the contact and line components) may serve to diagnose abnormal heating associated with defects in the electrical contacts. The use of ultraviolet-sensitive detectors or silicone CCD cameras may also be suggested for the diagnosis of abnormal spark intensity in contact with the catenary.

The Detection of smoke and / or gases and vapors can be achieved by to properly install analyzers and detectors, e.g. as proposed in Patent Document [004]. To this Way you can the information about the possible Smoke emission from one or more vehicles to supplement the Fire diagnostic method which are used here in section 0. The detection of dangerous Gases and / or vapors however, can be used within the functional sets discussed in section 0 be to the system in the context of risk reduction to make more useful in the railway transport of certain dangerous goods.

Infrared- and ultraviolet fire detectors can also be integrated into the system be for more data for to receive the diagnosis of fire aboard certain rail vehicles. The applicant notes that such integration is an interesting one Option represents if no passive infrared sensor matrix or no scanners are installed on the SMI to the ones in section 0 above discussed While using recognition methods it would not lead to any significant improvement in performance if said methods of this invention already in the system implementation be used.

A another special feature that will be integrated into the system can, is the detection of radioactive sources, what special for metal waste loads on wagons leading to ironworks be brought [042], as well as for national and international security aspects related interesting with the smuggling of fissile material.

in the Generally, one can integrate further means of detection from faulty and dangerous Conditions in rail vehicles pass from the basic properties emanating from the method, and especially from the availability of specific data the vehicle models and information from the vehicle database and from the use of VCPO functions, thanks to them, a precise tuning of ground-based instruments with one or more parts of the vehicle is possible. Other general The benefits of such integration arise from the possibility of the integration and communication features and means (hardware and Software) discussed in Section 0 below use.

5.18 Data acquisition

As mentioned at several points in the text above, the measurement data acquisition by the system must be performed so that its exact time is directly or indirectly allocated to each measurement is used, and that a single time scale is used or that the measurement times, which are defined by different time scales, can be reconciled by a coincidence of the time scales. More specifically, the time associated with a measurement (with the required accuracy) should be the physical interaction on which the measurement is based (eg the time a laser pulse is delivered by an LDM or the average exposure time of a CCD or a thermal IR). Sensor against electromagnetic radiation).

The actual series of measurements requiring data acquisition clearly depends on the many options discussed above for the instrument and sensor types that are integrated into the system, as well as their number. The required accuracy of the measurement times is also variable for different instruments, mainly because of their bandwidth and their output characteristics, but it can generally be said that with respect to the channels of higher measurement rates, a typical time accuracy (in the sense of uncertainty regarding the difference between ranging from 10 ^-5 to 10 ^-4 s, and such values correspond to approximately 0.3 and 3 mm of uncertainty relative to the longitudinal position of a vehicle traveling at a speed of 120 km / h.

The Implementation of the system is very likely to be done using the commercially available Instruments and components that are continuous or separate Generate signals and / or digital outputs, and in accordance with the different standards or with their own patented standards. In particular, some of the data to be acquired (e.g., the signals Most wheel sensors) available as voltage or flow and correspond to an analog signal or a two-state output, without any synchronization with the measurement acquisition system takes place. Create other systems (e.g., certain CCD cameras) an analog signal or a digital record with a synchronization the true measuring time and the corresponding output time by Input and output triggers or clock signals are implemented. In some special cases (e.g. in the case of wheel sensors with two-status output) captures the actual Measuring the time related to an event, even if in practice the data acquisition technique through the regular reading and saving and / or Processing and saving the corresponding output implemented becomes.

The Data acquisition equipment and the appropriate software can Developed by an experienced designer, as the applicant foresees no particular difficulty, and because a number of Systems so far in various fields of engineering and Research Science have been implemented where stricter requirements to the measuring accuracy, the time accuracy, the number of measuring channels and the Throughput of the measured data are provided. The applicant has it however, preferred to include the following text in this section, to show that there are several options for development and implementation the data collection for the system possible are, if one is easily available used industrial components and systems.

A very attractive solution for the Implementation of data collection in the system is the use a "crate" the VME family [971] or VME "Crate" collections, thanks to availability a big one Selection of motherboards, insulation inserts and plates, in real time operating systems and appropriate software development tools. VME systems are also almost a natural one Choice "for the corresponding Industrial area and they would be in the special case, especially attractive because of the bus-operative Properties and availability of VME bus lines for time measurement and synchronization purposes.

22 shows some typical solutions that can be used to meet the appropriate timing requirements when implementing the data acquisition into the system using a number of independent data acquisition units that do not provide standard solutions (hardware and / or software) for such requirements Offer. In general are 780 . 786 . 789 and 799 Data acquisition units, which typically include at least one CPU disk and one or more specialized I / O loops or disks. Such data acquisition units are connected to a network (eg, a Fast Ethernet LAN) to perform various functions, and above all to transmit the acquired data, but it is understood that in this discussion such a network would not be capable of implementing the exact synchronization the corresponding clocks of the different units is used. The 780 works at least as a master timekeeping unit in data collection, but it can make its own series of digital and analog inputs 782 and a series of digital and analog outputs 783 have directly connected to one or more meters and sensors. The unit 780 could be a VME crate with one or more CPU disks and a programmable digital I / O toughen ler / Uhr / Zeitkarte [971] or an industrial PC based on an Intel chip [972] with a multifunctional I / O card, such as certain I / O cards manufactured by National Instruments Corporation of Austin, Texas, USA [969, 970]. The digital outputs containing the synchronization signals of the connections 787 . 793 and 805 are generated by programmable counter circuits or by other circuits operating within the unit 780 can be programmed. The output and input signals of the unit 780 have a timekeeping on a single "master clock" 781 based. The units 786 . 789 and 799 may be different industrial PCs, as mentioned above, having the necessary digital and / or analog I / O cards and data exchange ports (eg, USB II or IEEE 1394) as required by the corresponding peripheral meters.

The unit 786 illustrates the detection of an analog continuous signal (without synchronization of the measuring sensor) from a sensor 784 via a signal cable 785 , You can create a data acquisition card within the unit 786 use, so the data conversion externally by a signal from the dial gauge 788 is triggered by the connection 786 from the unit 780 is delivered. Registered this way 786 not directly all times of the measurements from the sensor 784 but only the times of the measurements, in the appropriate format within the unit 780 which are the measurement trigger signal 788 generated. A value for each measurement date is stored and / or transmitted through the unit through the LAN 786 and the association of the measurements and the times takes place asynchronously.

The connection scheme of unit 786 can also be used in the case where the device 784 Event signals produced by a two-state transition and these transition times must be measured (eg in the case of certain types of wheel sensors). In this case, a counter circuit of a digital I / O card can be used to control the signal pulses 788 to count, and he is read out when the signal from device 784 his status changes.

The connection of the data acquisition unit 789 illustrates the case where the meter 792 is triggered externally, as in the case of a single-beam laser distance meter for the time of passing. Measurements are made by the signal 794 and the measurement is triggered by the signal connection 790 detected when a data acquisition trigger signal from the connection 791 Will be received. Also in this case, the data acquisition unit does not register time values and the times corresponding to the measurements may be out of the unit 780 be read.

The connections of unit 799 illustrate another, more complex case in which the measuring unit 800 (eg a linear CCD camera) a measurement trigger signal from the connection 803 accepted and the signal through connection 801 with the synchronization signal 802 generated for the data acquisition. The connection 805 can in this case be considered an external trigger or as a clock for the timing of the trip signals to unit 800 be used, and the connection 806 can be used as a digital enable / disable signal to start or stop the series of measurements.

Special solutions can be provided for special cases. For example, a Zoller + Fröhlich laser rangefinder scanner [961] may provide a measure trigger signal from unit 780 and asynchronously generate the distance values along with the corresponding rotary encoder data through a fast IEEE 1394 serial data connection to an industrial PC having such a type of data port.

As in the comment above 4 is discussed, the measurement rate for some of the sensors and instruments that are part of the system is defined before data acquisition begins, and it depends on the speed of the incoming train, especially a waste of data storage space and processing resources avoid when the train speed does not require the use of the fastest available measurement rates. A reduction or increase in measurement rates while passing the train of the SMI may be implemented, if deemed necessary.

In general, it is clear to those skilled in the art that data collection can be implemented by a variety of means, and that the discussion above is to 22 represents only a part of the possible reasonable solutions. It is also clear, though not discussed in this text, that signal conditioning, isolation, grounding, etc. must comply with applicable electrical safety regulations and avoid interference that can be caused by strong electromagnetic fields in the environment the railway line at the SMI can occur.

5.19 calibrations

One Main aspect of the calibration for the system involves the calculations used, as discussed above which are measurements taken by instruments installed on the track carried out be associated with elements that relate to the passing Vehicles are located. When discussing this topic, you can have different consider important aspects, like the "geometric Calibration "of the instruments themselves, the geometric calibration in connection with the installation of the Instruments and the possible Shift or change the positions and orientations of the instruments relative to the Railhead.

The Geometric calibration of the instruments can be very handy offline carried out be, e.g. in a laboratory, from the instruments like cameras, IR imaging devices, IR scanner or VLDS, for the different measuring directions with different pixels or the position of one or more scanning elements (for example, by an angle encoder). This type of calibration can on the basis of a (polar or Cartesian) with the instrument integral coordinate system, so that the installation the instruments defines an ordinary "rotation-transmission" of the entire set of instrument measurement beams. This offline calibration is generally for instruments with a single fixed measuring beam not necessary. additionally it can do this for all appropriate instruments required or recommended be, offline beam profiles and the corresponding accuracy parameters to eat. The ones mentioned above actual Offline calibration processes can also the adjustment of mechanical and optical components to align the optics and to optimize the measurement performance. Most of the appropriate tools for the system are such that the instrument-specific geometric calibration is performed once (e.g., "at work" or in a laboratory, before installation). Your after-attitude may be after a certain Be required after installation (depending on instrument type), e.g. due to alignment changes from internal components of the instrument due to persistent vibration or because of a random one mechanical shocks.

A second type of calibration, involving all of the optical (VIS, NIR and IR) instruments mentioned above, is associated with the determination of the position of the coordinate systems based on the instrument, or directly with their measuring beams, and is done in an ordinary static coordinate system such as above explained C _GB . This calibration can be performed after installation of the instruments at the SMI and generally requires some common accessories such as hyperstatic three-dimensional frames positioned along the track and these should preferably be mounted on an appropriate railcar. The particular and important case of electromagnetic wheel sensors has been discussed above and is in practice to associate a relative longitudinal position along the rails with the corresponding "travel time" of the sensors, ie a wheel set whose displacement is precisely measured over time.

Another aspect related to the calibration relates to the slow shift of the position of the rail head relative to the instruments along the track, caused by rail head wear, ballast deformation, and the displacement of the instrument position due to small deformations and displacements of their supporting structures attributed to earth movements and temperature changes. This aspect can be accounted for in various ways, mainly depending on particular decisions in the mathematical formulas of the coordinate conversion process discussed above, which instruments are installed, and for which instruments data processes are implemented in the system. One important choice concerns the definition of the C _GB coordinate system, which may either be integral with the rails or integral with the supporting structures of the instruments. Such an alternative leads to a different mathematical formulation for the consideration of said shift. The displacement of rails relative to the group of supporting structures of the instruments may, if desired, be measured by the ad hoc installation of sensors, such as optical distance sensors integral with the instrument structures of the system, and the measurement relates to Distance from a part of the rails or from some mechanical elements mounted on the rail. Another way of achieving the same results with sufficient accuracy in practice is to process, as explained above, the wheel images or measurements made by at least two laser distance meters positioned to provide the same observe the lower outer surface of the passing wheels. Another way to monitor the positional shift of the rails is to measure the VLDS, which, if available, allows for the determination of the exact average of the time intervals when no train is present in the SMI. In general, and given that this question as a whole relates to the rail shift versus the instruments, irrespective of "Earth position", different solutions can be based on the measurements made by the optical instruments of the system (VIS and IR) of geometric elements located in the area next to the rails or for this purpose (eg fixed panels, which can be positioned through the Cameras are observed).

The change the rail head position through maintenance, and especially by rail grinding and tamping and leveling the rails is different than changing the railhead position, which was mentioned above, since it is not a slow one Displacement is and is due to certain and known maintenance interventions is caused. A re-adjustment of the system can after such special maintenance may be required, also because they are the Repositioning the system sensors installed on the track are necessary (e.g., wheel sensors and infrared scanners for axle-related elements).

A volatile change the corresponding position of the rail head relative to the Instruments as a result of the forces that of vehicle wheels exercised strictly speaking, there is no aspect related to the calibration, but her discussion has a lot to do with the above explanation about slow Shifts and sudden changes in the railhead position in relation to the instruments in common. Such volatile changes can taken into account if maximum accuracy required is (through some of the above for the displacement of the rail head position mentioned methods).

in the However, it is generally for Professionals in this field realize that the application of the above most accurate methods for the determination of the functions of Ω and Γ very good performances of different Detection functions above for the defects and the dangerous ones Conditions were explained enable, without a significant impact on the changes (displacement, sudden change and fleeting Change) to have the rail head position.

A Series of tests for "integrity monitoring" in relation to the Shift or change the relative position and orientation of instruments vis-à-vis the Rails should be checked regularly the system software performed be, and such tests can practically in the above mentioned Re-calibration methods are integrated.

One Another aspect of the calibration is timekeeping. Some of the data collection processes can indeed virtually constant and non-negligible Latency or delays in the acquisition of measurement data (e.g., late delays or the timing line). These timing parameters should be appropriate for the different ones Cases considered become.

Other Calibrations refer to scalar quantities directly through some the system instruments and especially by passive infrared sensors for thermal Emission measurements and be performed by laser distance meter. The calibration of thermal emission based temperature measurements became previously mentioned in this document and she should done with the help of known solutions, depending on the instrument type and depending on how far the instrument automatic calibration functions fulfilled. Laser distance meters generally do not require re-calibration for very long periods of time, with the exception of the measurement changes due to the displacement of the position or orientation of the instruments, that were mentioned above.

It is of course possible that the re-calibrations associated with rail positional shift, instrument position and orientation displacement, distance measurement, wheel sensors and thermal emission measurements may be performed by a maintenance crew based on a programmed or adaptive maintenance schedule , Applicant once again emphasizes that this document details the method and system, leaving a number of options regarding system hardware and software methods. The above explanations seem to indicate that such options are related to the whole question of the methods for the "geometric calibrations" and the "geometric re-calibrations" that should be dealt with at the moment of development of the system implementation. The box 243 in 3 generally refers to the configuration and calibration functions, but those skilled in the art will recognize that these functions are implemented by a variety of different software modules and a portion of them with other boxes in 3 can be associated.

5.20 Vehicle database

Of the above text of this document explains clearly that the "vehicle database" is a fundamental component of the system is because of a set of critical data and information to be retrieved from it in the implementation of the procedure Related to the vehicle models or to the models of execution vehicle components (e.g., axles and bogies that may be used for more than a vehicle model can be the same). It is also stated above that the amount and complexity of the data and information in such a database (or series of databases) largely variable and for Different vehicles and component models each different could be, dependent from the installed instruments, the processing method used the records and the level of detail in the application of different functions desired within the procedure is.

The Structure and implementation of the vehicle database should be included the development of the system implementation of professionals in this Be defined area. The applicant wishes to mention that an object related Database [070, 071] is a particularly elegant and effective choice for implementation could be the vehicle database, out for Professionals in this area obvious reasons.

Also if some copies of the vehicle database on different servers in a WAN (Wide Area Network) may be present the applicant intends to make a copy within each individual system installation to use for availability system installations and to prevent network and database access infrastructures with guaranteed minimal services needed be slowing down some critical system processes to prevent.

The Vehicle database should be from one or more organizations according to one reasonable defined technical plan. The To be copies of the vehicle database in the system installations, preferably by an automated maintenance application, through Updated one of the network connections below.

The Enter and manage the data and information stored in the vehicle database for the Vehicle models can be stored by software applications are supported with certain features that reduce the workload reduced and the reliability the management of the vehicle database improved. A home solution that To facilitate this administration is the use of a three-dimensional CAD (Computer Aided Design) application especially with respect to the Definition of the model-specific geometric described above surfaces, e.g. the TESD and HTDS elements. The CAD data files for the vehicle models, together with the data and information from the measurements and from the vehicle database can also used to be useful create graphs that diagnosed the Defective and dangerous Conditions. For example, it is possible on the computer screen in a railway control center a three-dimensional view of a Imaging vehicle in the surface pixels colors accordingly assigned to the temperature or position relative to the Ladeprofilumhüllung become.

Also if in the vehicle database only data and information about the execution model necessary, To implement the system can also provide certain information for individual Vehicles in relation to an execution model stored in the database (e.g., for certain fleets) related to with some process's processes or with value-added functions stand on the integration of the system into one or more (other) systems can be based.

5.21 Communication & Integration with external systems

Local system components are defined herein as the system components (hardware and software) installed at the SMI or close to the SMI, which represent a system installation. Thus, the local system components are those in the dashed area of 2 are included (plus the interconnections between these components). Remote system components (hardware and software) typically (but not necessarily) belong to more than one system and are located at a variable and as large a distance as possible from the individual system installation (s). As used herein, the term "external system" refers to the signaling and security systems of the railway or to various railway information systems that can communicate with or be integrated (directly or indirectly) with one or more system installations or with remote system components. 23 generally shows the communication and integration between remote system components, local system components, and external systems.

box 815 shows some important local system components of a system installation. The boxes 810 to 813 and box 821 show a number of data processing units that are connected in a local area network, eg by a fast Ethernet or a Gigabit Ethernet LAN with one or more switching and socket units 814 , For example, you can 811 and 812 two data processing units, which are mainly for data acquisition while 813 a data processing unit used to run some of the software applications described above.

The grouping box 834 , including the remote system components, includes some data processing units (from 827 to 830 ) and network components (eg a Fast Ethernet switch 833 ), which are installed at a remote location and form a System Remote Management Center (here "SRMC"). A key feature of an SRMC is the monitoring of a number of system installations to detect possible faults or malfunctions and remedy them through remote management applications or by scheduling a maintenance crew intervention. Such monitoring may be based on a "center request" alert scheme and / or sending appropriate messages from the system installations to the SRMC, if something abnormal is automatically detected by the diagnostics functions, or sending regular status indications from the system installations to the SRMC SRMC, where the delay or absence of such messages from the system installations is interpreted by the SRMC as a symptom of malfunction. Another important SRMC feature is the updating of the software and database of system installation data files. Such updates are possible by applying storage means to the installation, but if possible they should be carried out using one of the means of communication given below. Software adjustments and updates may be made in an SRMC or elsewhere, as well as updating the contents of the vehicle database, and then sent to one or more SRMCs for distribution to the system installations. Some of the software updates will be implemented according to update plans (introducing new features, improving code performance, etc.), while other updates may result from the diagnosis of code "bugs" or identifiable malfunctions, and they should be completed within a short time be executed, if possible by the SRMC. The communication of the system installations with an SRMC is also very important in connection with the data transfer from a system installation to a missed identification of a vehicle and the generation of any alarm (in this way the responsible engineers can directly check the data of the system installations to determine the origin of the system Failed to verify any identifications or alarms and, if necessary, propose or implement one or more changes in the system installations or the content of the vehicle database The records from a system installation to an SRMC will also be submitted for off-line work, as described above More than one SRMC could work for a number of system installations based on different layers, different purposes and / or redundancy it fulfills several functions related to communication with external systems, as explained below.

in the Generally, the Applicant observes that the TCP / IP ("Internet") protocol for most explained in this section communication transmissions could be used with some obvious exceptions like the two-state signal lines to a signaling relay (see below). Other network protocols can also be used in this system implementation, what for professionals in this area is obvious.

The element 824 represents the Internet network, which, with the appropriate safeguards, can serve as a primary and practical means of connecting the system installation to one or more SRMCs. The use of the solution VPN [072, 073] is an example of such a security device. Accordingly, the box represents 819 in 23 a hardware "VPN customer" like a Cisco System 3002 VPN unit [974] while the box 832 Displays a "VPN Server" in an SRMC, ie a Cisco System 3000 VPN unit [973]. Of course, other means (firewalls, surveillance systems, password management, operator personal identification devices, etc.) should be chosen by an experienced engineer to best suit system implementation or upgrade activities. The boxes 820 and 831 represent the units for Internet access, for example DLS modems / routers. Private networks, leased lines or virtual private connections made by telecommunications technicians are, of course, some of the alternative options for using the Internet. Satellite links may be an option to establish a connection between the system installation and an SRMC, especially with regard to monitoring the system installation when the primary network connection means is not available. As an example of a cheap redundant satellite-based connection that serves as a monitoring backup sent the element 818 an access unit such as [976] for the global satellite communications system "ORBCOMM" [975] represented by Element 825 , This solution provides a simple interface with an SRMC by sending the appropriate messages over the Internet based on the standard e-mail protocol.

unit 822 stands for a wireless modem that could be installed in a system installation to provide a wireless backup connection in addition to or instead of the satellite-based connection. Using a GSM or GPRS or UMTS or other wireless network 826 would of course imply communication with a corresponding ISP, if, as in 23 presented access to an SRMC over the Internet. However, direct "dial-up" connections can be made between 815 and 834 which are based on wireless modems on both sides. The boxing group 842 that the boxes of 835 to 841 that covers the boxes 827 to 833 can match another SRMC like 834 or an external information system such as a rolling stock management system or a maintenance information system, as further explained at the end of this section.

box 845 stands for a signaling and security system of the railway company to which the system installation sends, in the form of messages or signals, certain information and / or alarms originating from the detection functions of defects and dangerous conditions discussed above. In the example of 23 manages the data processing unit 817 the news and signals between 815 and 845 , unit 817 Creates the connection (eg via a standard serial connection such as RS232) to the data processing unit 821 and not to the other data processing units used by the local area network 814 is implemented. This solution allows the exchange of only certain series of digital messages between the units 817 and 821 to ensure the safety of the unit 817 to certify running software, especially if the connection between 815 and 845 a digital network is (eg based on TCP / IP in an ATM network).

One of the simplest ways for the system to send security alarms to the signaling and security systems of the railway company is to use 2-status signals via a relay. box 853 represents one or more relays and possibly some subordinate electronic circuits to the corresponding unit 844 to signal an alarm that is integrated into the signaling and security system of the railway company. Different signal lines can be used to indicate various alarms, such as "fire on board", "hot runner", "out of gauge line", etc., also because different signaling can be performed. The system's hazard reduction mission generally does not require that the protocol for this communication meet the safety integrity principles for a fail-safe system connection, and thus various protocols based on NO and NC relay contacts can be proposed (provided they do not de-classify the SIL stage the signaling system).

Additional signal lines may also be integrated, and some of them should allow signals from 844 at 853 to implement the signaling integrity protocols (eg, a request from 844 at 853 to activate a confirmation signal from 853 at 844 to check the availability of the system installation and the corresponding signal cable). The use of this interface technique between 815 and 845 is suitable for signals related to the stopping of a train or its detour when dangerous defects or conditions are detected. Creating interfaces via relay signals is in any case the "most natural" way of creating an interface between the system and multiple "traditional" signaling and security systems.

The use of other communication techniques between 815 and 845 as well as a digital communication line through the units 823 and 843 Through this, in addition to implementing the same basic signaling functions mentioned above for the relay interface, it enables a number of more complex functional relationships for the interaction of the system with the signaling and security systems of the railway company. A first example of these other functions is the exchange of messages, as discussed above in Section 0 for the occurrence of loading profile violations or the acceptability of large-load on-carriage transit depending on speed restrictions. Other such functions relate to the transmission of data and information from a system installation to a railway control center in cases where the system has identified a potential hazard that needs to be validated by an operator in the control center (eg a loose wagon or sheet) Violation of a boundary line profile that is a consequence could be a load shift, but at this stage is not yet such that the alarm triggering conditions would be violated if a boundary line profile were violated). When an operator's judgment is needed, the system may send a combination of data and information that may include some mappings of the corresponding vehicles and the corresponding hazard (eg, created by the linear VIS cameras and / or linear IR sensors and / or VLDM instruments installed on the SMI, as explained below).

Railway signaling professionals will know that the current increasing use of GSM-R combined with ERTMS / ETCS is another way of communicating and integrating between the system and the railway signaling and security systems, especially for communication with train systems and railway operators for integration with the signaling and security infrastructure (eg in connection with the speeding of the transport of wagons with very wide loads). Considerations on safety and economy (eg [009, 023 and 026]) lead to the realization that when a train formation due to the detection of defects or dangerous conditions at a manned application site on the rail side (or at a suitable location on a rail track, to which a service crew could be sent) the appropriate information should be available on the spot, specifically the information on the vehicle (s) that need to be checked and, if applicable, the part (s) of such a vehicle / such vehicles (eg which axle box). The group box 852 corresponds to a set of remote components, including remote components of the system, associated with the management of train formation with vehicles at a station or marshalling yard or on a suitable track section upon detection by the system of defects or hazardous conditions on one or more vehicles Train composition. The communication between 815 and 852 can be done by a variety of alternative transmission technologies, for example by the use of special glass fibers or a copper-twisted cable pair between two suitable communication units 816 and 850 (eg two modems / routers). In the overview of the survey in 23 is a PC 848 with a system installation 815 via a local area network, eg an Ethernet LAN based on the Ethernet switching unit 851 , connected. In addition, a "handheld computer" or a "portable computer with touchscreen" 846 with an adequate communication device 847 be equipped to a wireless data connection via 849 to the LAN of Box 852 to offer. A local outdoor wireless network could be, for example, a FHSS (Frequency Hopping Spread Spectrum) w-LAN implemented with the necessary components [977] from Alvarion of Tel Aviv, Israel, and a handheld computer for this the corresponding PCMCIA plate [978]. The advantage of this particular w-LAN solution for this application is the connection of the handset with a data transfer rate of several Mbps over a distance of several hundreds of meters or up to kilometers, which offers a very high data security and a remarkable immunity to interference. When a crew uses a handset, it can obtain information about the location of the detected defect (s) or the dangerous condition (s) of the stopped or diverted train. Operation instructions can also be given to the crew via the handset and a robust software interface, similar to a checklist, makes it easier for the crew to intervene and contributes to the corresponding quality assurance. The unit 848 can be manned by an engineer who can track and communicate with the crew's activity and, if desired, can also activate the verbal communication feature of the above-mentioned FHSS w-LAN. Pictures of the system can be attached to the units 846 and 848 can be considered, and, if desired, the unit 846 with a handheld device (built-in or connected to the camera) retrieve images of the crew mission (eg a shifted cargo) and send them to the unit 848 send. However, communication with the laptop of the service crew can be done in other ways, for example via a wireless telephone network. It is clear to those skilled in the art that the functions associated with crew intervention can be supplemented to include also railway company control centers and other rail information systems. The implementation of such a service crew deployment system also provides an effective and high-quality means of transmitting to SRCM the actual observations of the crew, and these are very important to the evaluation and improvement of system performance.

However, the verification of a possible defect or dangerous condition may alternatively be performed by a crew composed of railway personnel operating on the train concerned. In order to notify such a crew of the detected defects and / or dangerous conditions, the same information discussed above with the use of a ground-based crew may be communicated on board the train with other communication means, especially via a GSM-R network or other wireless communication systems. The information and data may be communicated to the train crew in the form of vocal messages and / or digital messages to a computing device with an appropriate screen (possibly with drawings and / or images sent) and / or in the form of a fax. In addition, "information pillars" may be positioned along the mainline or on a safe rail section where a train would be stopped, and these columns could be connected to the appropriate system installation or to a railroad system. As noted above, for most of the identified vehicles, the system recognizes a unique vehicle identification code in addition to the vehicle execution model. The use of all identification information allows the system to generate important information for the rolling stock maintenance management systems and the logistics systems. The detection of abnormal conditions or tolerable defects that do not require stopping or diverting the train may result in the transmission of a message to a maintenance information system in which the diagnostic information is associated with the unique vehicle identification to be used by the vehicle be considered different purpose, so for example, when advancing the next maintenance. The subsystem of the units in the box 853 can be integrated into the maintenance systems of rail vehicles, for example, to send a field of application of a crew and / or to plan a specific maintenance intervention for a vehicle and / or to get the necessary spare parts. The integration of system installations with the SRCM (s) and maintenance-specific information systems can include a number of value-added functions, such as automatic generation of large data sets to improve maintenance planning, optimize condition-based maintenance, and improve system software performance. These interfaces with logistics-related information systems can be implemented via a database server, which records the passage of a particular vehicle to a system installation. Such a server could be interviewed by various logistics-related information systems which ideally have Internet access. Alternatively, an SRMC may send messages to an information system of the respective fleet owner containing the location of the installation, the time of passing and the unique vehicle codes.

5.22 Software implementation

As Anyone skilled in these techniques will require implementation of the process and the system a considerable amount of software development, compared to other major development efforts (executive design and construction infrastructure elements, instrument housings and their installation, detailed design and setup of the data acquisition hardware etc). The applicant holds, however The development of the system software is not critical, as most Special Critical Algorithms above in the detailed Description were presented. Consequently, a suitable Team of qualified software engineers the necessary ones Developments based on this document, by public Information and the current state of the art in relation to the development of multi-processor high performance computing applications To run. The experts are clear that a number of alternatives for the Implementation of the system software exist, especially what the detailed Design techniques, the programming language, the software development kits and the operating systems concerns. The following text in this section is limited therefore, to a few remarks, which for professionals probably quite not necessary are.

Developing a software application in C or C ++ language in a real-time software environment (eg, RTUnixPro [979]) is considered by the applicant to be a highly recommended choice for the software modules of the system, such as the box 218 in 4 (especially if the corresponding data collection functions are implemented by a VME system) and those that use the boxes 821 and 817 in 23 as explained in Section 0 on the interfaces between a system installation and a railway signaling and security system.

Choosing, based on the information in Section 0, to structure "heavy" software functions for large calculations as a series of software modules that perform asynchronously certain tasks for a particular vehicle of a train, and as multiple instances on a series of calculation units of a network or a collection is compatible with another software design solution, and especially with the use of the monitoring application 231 whose tasks include the allocation of computational tasks to a set of software modules that run asynchronously and in parallel. The use of a typical real-time software environment is generally unnecessary for asynchronous modules that can run under operating systems that are different from that used for the computing units running time-critical applications. In general, ease-of-maintenance, modularity, and portability criteria are very important for the design and development of system software, as well as for most professional software systems that need to provide a life cycle that spans many years.

5.23 Installation, accommodation and protection of the installations

A first important aspect related to the installation of devices is the mechanical stability of the installed sensors and instruments, especially the (active and passive) optical elements, which are positioned on the rail sides as explained above. The system here does not require a special type of support structure for such sensors and instruments (eg a signal bridge [002] or arched structure [003] or the walls and ceiling of a tunnel [066]) and therefore engineers responsible for the design can use the Select the type of structure that you consider appropriate (eg a group of lattice masts or a series of interconnected metal lattice masts along the rail side or structures of composite concrete and metal girders), provided that it is ensured that the vibrations and the positional displacement and orientation the sensors and instruments are within the limits specified for the intended use of the system. In general, it is not difficult to keep linear oscillatory displacements of optical sensors and instruments within the desired limits (usually a few millimeters), while angular fluctuations can be more critical. In any case, the vibration limits are given by the required accuracy of a specific measurement together with the corresponding measurement geometry. For example, in the case of a very fast laser rangefinder, as in 13a 4 mm from the axis of oscillation of the instrument, place a fluctuation distance of a width of ± 5 milliradians (eg, approximately ± 0.3 degrees) to a vibration of a width of approximately ± 20 millimeters of the measurement point orthogonal to the measurement beam.

The Design of supporting structures for the instruments should also slope of the train, the wind and the effect of temperature changes to consider the structure. If applicable, the deformation effects should be related with the seasonal occurrence of "permafrost" (freezing and thawing) balanced, e.g. by having a special reinforced concrete foundation for the entire SMI installation.

The Question of mechanical stability must of course also in the choice and conception of fixing the instruments to the o.g. supporting structures and the evaluation of the structures the instruments themselves be (for example, it can actually be that linear Vibration by a passing train passing through the supporting structure the instruments is forwarded, an angular vibration in the Arrangement of the instruments and its fastening components causes).

Of course everyone has to Hardware elements mounted on the SMI comply with applicable safety regulations comply with and be installed and maintained to prevent accidents be, e.g. by falling or moving physical elements or by an electrical overhead line of the railway (including the accidents due to the displacement or failure of power lines of the Railway company caused). An addition to be considered Aspect is the protection of the exposed optical surfaces of some Instruments against weather influences (Raindrops, drizzle, snowflakes, etc.) and dust, greasy splashes and other particles and / or small objects (possibly brought by the tension suspension and / or by wind), these optical surfaces change and affect the proper functioning of the system can. As above in some sections about the instruments for mentions the system implementation different methods have been and are already known applied (automatic closing a protective lid or flap when an instrument is stationary, the rinse the exposed optical surfaces with clean air, heating etc.) and the for System implementation responsible engineers will be theirs Opinion to choose appropriate combinations of such methods adequate protection of the exposed optical surfaces of the ensure appropriate instruments of the system.

Of course, sun and / or other intense light sources and / or sources of infrared radiation can be a problem for system operations as they affect the measurements or availability of optical and / or infrared instruments. However, it is envisaged that the system may be installed anywhere along the rail track without precluding locations having particular geographic orientations above the rails or in which intense light sources or sources of thermal radiation are at a critical distance from and in critical Direction related to the SMI. One of the techniques with which such potential problems of sun and artificial light and / or sources of thermal radiation can be prevented is the use of special attachments. In the case of linear imaging devices (VIS, NIR, and IR) and those instruments with which an angle scan is performed (eg, scanning laser rangefinder), the use of an attachment is similar to 541 and 548 (in 13a and 13b ), along with a reasonable elevation angle The range of the rays of the optical instrument (alone or by associated means) is a very effective solution for preventing both the potential problems associated with sunlight, intense light and / or thermal radiation, as well as weathering and airborne particles. Another solution to the potential problems associated with sun and / or artificial light rays is the installation of shielding plates on the supporting structure of the instruments, with the panels mounted on the opposite side of the rail from the corresponding optical device or infrared instrument. These panels do not need to have solid surfaces and the use of panels with small openings is recommended to reduce the wind pressure that would otherwise require costly reinforcement of the supporting structure of the instruments and panels. However, these panels can be installed on one or more structures, which need not be the supporting structures of the instruments. The preferred use of linear imaging devices in the system, in addition to the advantages discussed above, also has the advantage of allowing the use of relatively narrow shielding panels, rather than the wider panels typically required for shielding two-dimensional imaging devices. Installation of the system in a railway tunnel [066] or under a wide bridge or under an ad hoc protection tunnel (to protect and possibly carry some of the system equipment at the SMI) is of course possible but not required for the system , In particular, the Applicant believes that the construction of a special tunnel-like structure (and possibly a support for the instruments) would probably be unfavorable in purely economic terms.

5.24 Examples of System Configurations

Some Examples of system configurations are discussed in this section presented briefly, especially related to various possible combinations of sensors and instruments installed on the SMI.

5.24.1 First example a system configuration

This first example relates to a relatively complete configuration to support all of the important recognition functions outlined above and to provide high performance in terms of detection sensitivity of all major defects and / or hazardous conditions at a low false alarm rate to reach. The reference speed range of the vehicle (at the SMI) is 35 to 120 km / h for the application of the complete set of recognition functions, and covers a range of 20 to 160 km / h for the vehicle identification function and the detection function of fire and heating of a vehicle body. There is a tolerance of 5% at the top and bottom of these speed ranges. Detection of oncoming train formation and arrival time prediction at the SMI of this train formation is based on two pairs of Honeywell RDS80001-H [950] wheel pairs, each pair along the same rail with a short distance between the two sensors of a pair (eg 1000 to 3000 mm), and the pairs in two positions like 206 and 207 with intervals 210 and 212 installed by the SMI of about 170 meters. 24 shows a simplified bird's-eye view of the sensors and instruments installed on the SMI for this initial configuration. The little black square 884 represents a high bandwidth electromagnetic wheel sensor, especially in this example a variable resistance sensor (VRS) [951] from Invensys Sensors Systems / Electro Corporation currently part of Honeywell). Five pairs of these sensors are installed on the rails in this example to provide the time data of the wheel transit to calculate the LDF function with high accuracy, even in the case of low braking speed rail vehicles. The distance between the wheel sensor pairs at the SMI is between 4 and 5 meters. All VRS and RDS80001-H sensors are clamped to or similar to the rails that are similar to or similar to Honeywell's RDS-CL-01 "Underrail Clamp", eliminating the need for rail drilling and correcting the position along the rail power. Two series of linear cameras 870 and 876 are similar to a geometry in the 10a and 10b installed, in this example they are Eclipse EC-11 cameras [956] from DALSA. The "periscope" configuration with the 2048 × 96 (TDI) pixel resolution and a 64.1 kHz max. linear sampling rate is used for the side cameras, while the version with a max. 17.4 kHz linear sampling rate is used for the down-facing cameras in a position like 449 and 450 are installed. The angle 464 between the rail direction and the sight planes 861 and 879 all cameras in this example is almost 90 degrees. The illumination for these cameras is produced with appropriate geometry (see section 0 above) by suitable fluorescent tubes or an LED array (preferably with 90% of their power in the wavelength interval between 650 and 850 ).

Four VLDS instruments 860 . 869 . 875 and 881 are similar to the SMI with a positioning and orientation geometry similar to those in the 13a and 13b Installed. Especially for this example the VLDS from the product line "Profiler" / "IVAR" [961] (mentioned in section 0 above) were chosen by the company Zoller & Froehlich GmbH. A version with a maximum unique range of about 25 meters and a measurement rate of up to 625 kHz is used, with a special housing with a shield and a cap similar to 541 and 548 in the 13a and 13b , Only a fraction of the 360-degree continuous "vertical scan range" is used for the measurements, and thus "near-real-time transmission" of the measurement data through the IEEE 1394 interface of these instruments is possible and saturation of the instrument's "buffer memory" becomes possible this measurement rate prevents (without the instruments themselves subsequently determining the average of the measurement data).

Four linear infrared imaging devices 862 . 872 . 871 and 877 are installed at the SMI. These IR imaging devices are model ISI (Plymouth, MN, USA) "Model IR 1000" [068], discussed in Section 0 above. The two units 862 and 872 are similar to unity 650 in 17a or 638 of 17b installed while the units 871 and 877 similar to unity 760 in 21a or 769 in 21b are installed.

Two fast laser distance meters 873 (the small compact triangle symbolizes both instruments) are mounted in a configuration similar to that in the 8a and 8b shown, in such a way that the two different wheel profiles with their laser scanning 358 measured on the wheel side surface at two different average heights above the rolling surface. Instruments from the Optocator ^TM [952, 954] range from LMI Selcom are used, specifically the Optocator ^™ 2008-180 / 390-B (part # 813214), a laser rangefinder with a range of 180mm, an installation distance of 390mm and a sampling rate of 62.5 kHz with a bandwidth of kHz.

The "threshold with instrument" 867 and the 6 other such thresholds in the group 864 are equipped with sensors that measure the wheel sets and detect defects in the wheel profile (wheel deformations, etc.) and correspond to the aforementioned "MULTIRAIL ^® WheelScan" system from Schenck [966]. The position of such instrument-equipped sleepers can be varied along the SMI without significant limitations (in this and other system configurations), except for the presence of special sleepers (not in this example) such as hollow sleepers used for infrared sensors, axle bearings, Scan wheels and brake components.

All instruments, with the exception of the instrument - equipped threshold sensors and the wheel sensors, are mounted on suitable steel pylons with reinforced concrete foundations (a mast for 860 , a mast for 881 , a mast for 875 . 876 and 877 , a wider mast for 869 . 870 . 871 . 872 and 873 and a small mast for 862 ). Narrow sunshades, as discussed above in Section 0, are installed on separate, simpler masts, if required because of the geographical orientation of the individual optical instruments. All optical instruments have an individual housing and an electro-mechanical or pneumatic flap to protect the optics when the instruments are stationary.

The data acquisition and data electronics units along with the power, communication and network units (and possibly with the production of compressed air and control components for the flow / pressure of pneumatically controlled valves) are in a weatherproof enclosure or canopy, about 3-4 meters away from the rails and closer 873 and 872 as on other instruments (to limit the overall length of the links and reduce the maximum length of the Optocator cables).

One meteorological sensor mast with instruments for measuring air temperature, relative humidity, wind speed and direction is mounted on a house or a shelter.

The following units are mounted inside (preferably in standard boxes for mounting on a 19 '' frame) of the house or shelter and in a Gigabit Ethernet network 798 via suitable cables and switch / hub units 814 connected:
a VME-based "master" unit for data collection 780 of wheel sensors, for the Optocator instruments (which use Selcom's VME board [953]), for the four models "1000 IR", infrared imaging devices for the instruments of the meteorological mast (via a serial data link with a meteorological data logger Dates;
an Intel Pentium data acquisition unit for each VLDS instrument;
an Intel Pentium data acquisition unit for each linear Eclipse EC-11 camera;
a variable number (depending on the maximum response time required) of data processing units, eg based [972] on an Intel Pentium or a Power PC microprocessor, mainly to run the software applications for the defect detection;
a data processing unit with redundant mass storage, which accommodates at least the vehicle database and the configuration of the system installation and the calibration databases;
a data processing unit for Schenck's "MULTIRAIL ^® WheelScan" system [966];
an Intel Pentium computing device, which primarily powers communication and signaling applications, and the signaling unit 817 and (optionally) with a backup communication interface (eg an ORBCOMM [976] 825 or a GSM 826 modem 822 ) connected is;
a router and a VPN hardware client as a single unit or separated into two units 819 and 820 for a secure internet connection 824 with one or more SRMCs;
a router / modem (as an option) for a high speed connection via a pair of drills or a fiber optic cable to a nearby station or other railway device 852 ;
one or more redundant data processing units.

The signaling unit of the system 817 is with the appropriate signaling and communication interfaces 853 and 823 connected. The data acquisition (taking into account the required time precision as explained above) is performed as explained in section 0. The real-time calculation of the approximate vehicle speed based on the Eclipse EC-11 linear cameras is also performed by the VME data acquisition unit. The VME data collection unit 780 has the RTUnixPro [979] real-time Unix operating system, while the other data processing unit has a different (ordinary) version of the Unix operating system.

5.24.2 Second example a system configuration

The second example is similar to the first, but there is only one of the two fast Optocator ^™ laser rangefinders and not four "Model 1000 IR" linear infrared imagers; these are replaced by scanning with IR photon sensors. "VAE-HOA / FO A400" scanners are used (mounted in a hollow sleeper) to detect defects and hazardous conditions in axle-related components, while similar instruments are used to measure the thermal emissions of railcar bodies and their loads.

5.24.3 Third example a system configuration

The third example is a configuration suitable for detecting defects in the boundary line profile, tread defects, and weight-dependent defects, but not for any type of defect and / or hazardous condition that can be detected by the heat release. The configuration of the sensors and instruments in this third example is similar to that of the first example, but without the two fast Optocator ^™ laser rangefinders and without the four "Model 1000 IR" linear IR imagers.

5.24.4 Fourth example a system configuration

The fourth example is a configuration suitable for detecting defects and / or dangerous conditions based on the measurement of heat dissipation, but it is not suitable for detecting defects in the boundary line profile (including profile infringement) Tread defects and weight-dependent defects. The configuration of the sensors and instruments in this third example is similar to that of the first example, yet without the two fast Optocator ^TM laser rangefinders and without the four linear IR imaging devices "Model 1000 IR", which by some laser distance meters for the time to be replaced by passing.

5.24.5 Fifth example a system configuration

The fifth example is a configuration suitable for detecting defects and hazardous conditions of axle-related components, tread defects and weight-dependent defects, but not suitable for detecting defects and / or hazardous conditions associated with the body (other than its body) weight). The configuration of the sensors and instruments in this fifth example is similar to that of the first example, but without the four VLDS instruments and without the two IR imaging devices for the railcar box and their charges. In addition, a small number of linear cameras are used with egg lower resolution (eg the Dalsa Piranha model 1024 pixels [955]) and equipped with synchronized LED pulse matrix illumination.

6. Glossary & References

6.1 Abbreviations and acronyms used in the text

• 3DD

  Three-Dimensional Data, i. the coordinates of a vehicle surface point in one on the Ground-based three-dimensional coordinate system and the corresponding Time, according to definition in section 0.

  BAC

  Beam Assignment Coefficient, according to definition in section 0.

  BID

  Buffers Information Data, information about the buffering of a particular vehicle model in the vehicle database, as defined in Section 0 with reference to 9 ,

  BIF

  Beam intersection Fraction, according to definition in section 0.

  BPD

  Buffers Profile Data, a record from a fast laser rangefinder positioned to measure from the rail side a profile of the vehicles at an appropriate height to detect vehicle buffers, as defined in Section 0 with reference to FIG 9 ,

  BWBTIS

  Bearings, Wheels and Brakes Thermal Infrared Sensors, as defined in Section 0.

  CVM

  Candidate Vehicle Model, as defined in Section 0 with reference to 9 ,

  CVML

  Candidate Vehicle Model List, as defined in Section 0 with reference to 9 ,

  CVMSD

  Candidate Vehicle Model Selection Dataset, as defined in Section 0 with reference to 9 ,

  ERTMS / ETCS

  European Rail Traffic Management System / European Train Control System

  External system

  A term used in This document uses a signaling and security system the railway company or various information systems of the railway company or to designate any other system that uses one or more multiple system installation (s) or with components of remote systems can communicate or be integrated into it, as defined in section 0.

  F (F1, F2, ..., F12)

  An element of a Car body, associated with a method of calculating one or several members for the determination of the VBPO function for one Car body, as defined in section 0.

  FHSS

  Frequency hopping Spread Spectrum, mentioned in section 0 concerns the wireless connection of portable computing devices for rail service crews.

  FLDM

  Fixed LDM, introduced in section 0 of this document.

  GSM-R

  Global System for Mobile Communications - Railways

  HLDS

  High-speed laser distance-metering Scanner, as defined in Section 0 of this document.

  HTDS

  Homogeneous thermal Diagnostics Surface, an area with the one representative Temperature within the data processing is associated with the the detection of axle related hazards, as defined in section 0.

  IDS

  Identification Data Set, as defined in Section 0 with reference to 9 ,

  IMA

  Imaging of possible marking areas, as defined in section 0 with reference to 9 ,

  IR

  Infrared; a far common abbreviation in physics and engineering.

  LDF

  "Longitudinal Displacement Function", the longitudinal displacement a vehicle over time along the rail route, such as defined in section 0 above.

  LDM

  Laser Distance Meter, introduced in Section 0 of this document.

  Local System Component

  A system component (Hardware and software) installed at the SMI or near the SMI is and is part of the local system installation, as defined in section 0.

  MSA

  Marking Searching Area, as defined in Section 0 with reference to 9 ,

  MTF

  Modulation transfer Function; a common abbreviation in the literature on the Engineering in the context of imaging sensors, imaging optics, Infrared imaging, image processing and target recognition.

  NETD

  Noise Equivalent Temperature Difference, the measurement of thermographic sensitivity [068]; a common abbreviation in infrared thermometry and the literature on thermography.

  NIR

  Near Infrared, i. electromagnetic radiation wavelength interval from the boundary from actinic red (about 750 nm) to about 3000 nm; a far common abbreviation in physics and engineering.

  OCR

  Optical Character Recognition; a common abbreviation in software technology.

  OCRO

  OCR Output, abbreviation used in this document, especially in Sections 0 and 0, with reference to 9 ,

  PV

  Previous Vehicle, as defined in Section 0 with reference to 9 ,

  Remote System Component

  A system component, related to one or more system installations and positioned at a variable and as possible huge Distance from the individual system installations, as defined in section 0.

  RPY

  Roll, pitch and yaw, according to definition in section 0.

  SMI

  "System Measurement Interval", the spatial longitudinal section along the railway track in which a passing rail vehicle is measured by the system, with the exception of the "train detection ranges" as described in section above 0. defined.

  SRMC

  System Remote Management Center, as defined in section 0

  TAM (TAM1, ..., TAM5)

  A temperature assignment Method, as defined in section 0 to be a representative Assign temperature to an element mounted on the axle.

  TEPP

  "Thermal emission data pre-processing algorithm ", a Algorithm (a group of algorithms) for computing some numerical values by data processing for a given data Subgroup of heat transfer data, that correspond to a TESD, as defined above in section 0.

  TESD

  "Thermal Emission Spatial Domgin", a definition (a group of definitions) to a spatial part of a car body to identify a TEPP algorithm for the measurements of heat releases is applied, as defined above in section 0.

  TESD opacity

  A number between 0 and 1, as defined in section 0.

  TESD transparency

  A number between 0 and 1, as defined in section 0.

  UV

  Unidentified Vehicle, used in this document, especially in sections 0 and 0, with reference to 9 ,

  WMPO

  Vehicle Body Position and Orientation, the special case of a VCPO function on the car body, as per definition in section 0.

  VBTHDM

  "Vehicle Body Thermal Hazards Diagnostic Method ", a method (a group of methods) that is used to make a diagnosis of dangerous Conditions in car bodies based on measurements of thermal infrared radiation create, as defined above in section 0.

  VCPO

  Vehicle Constituent Position and Orientation, used in 1 and in the corresponding comments, to denote a time function expressing the position and orientation of a "near-rigid" main component of a vehicle, and this function corresponds to the conversion of coordinates from one coordinate system to another, one of which is integral with the infrastructure and the other with the corresponding vehicle component.

  Vehicles Database

  A database that within the system serves to store data and information and retrieve those with a vehicle design model or with the construction model a vehicle component, such as a mounted axle or a Bogie, to be associated with more than one vehicle model belong can.

  VI

  Vehicle Identification, in this document for uses a process that is specific to a particular rail vehicle particular vehicle execution model which is encoded in the vehicle database and, possibly but not necessarily, even a distinctive code that the corresponding item (e.g., a serial number or a distinctive one Code within a fleet).

  VIS

  Visible; e.g. electromagnetic Radiation interval of actinic wavelengths, i. of about 400 nm up to about 750 nm; a shortcut, which is used in some areas of engineering.

  VLDS

  Very high-speed laser Distance-metering scanners, as defined in Section 0 of this document.

  VPN

  Virtual Private Network [072, 073], mentioned in section 0.

  WSD

  "Wheelsets Distances", the longitudinal distances between wheelsets, as defined above in section 0.

  WSI

  "Wheel Sensors Interval", the spatial longitudinal distance along a railway track between two outer elements of a series of wheel sensors installed on the SMI, as defined above in section 0.

  WTD

  Wheel Transit Time; used in this document, especially in section 0 to the data of wheel sensor measurements, i. the times when a particular wheel or wheel center of a wheel sensor or was detected by a Radsensorpaar.

  XSMI

  Part of an SMI, just as defined as the SMI itself, but neglected the wheel sensors, as defined above in section 0.

6.2 named as reference Patent documents

• 001 - DD 156450
• 002 - GB 2320971
• 003 - EP 1052606
• 004 - EP 1060766
• 005 - US 2818508
• 006 - DE 1031338
• 007 - DE 1082618
• 008 - US 3,095,171
• 009 - US 3226540
• 010 - US 3253140
• 011 - US 3646343
• 012 - US 4820057
• 013 - EP 0263217
• 014 - US 4878761
• 015 - US 5100243
• 016 - US 5201483
• 017 - US 5381700
• 018 - US 5331311
• 019 - FR 2752806
• 020 - US 3151827
• 021 - US 4441196
• 022 - US 6043774
• 023 - US 3721820
• 024 - US 4248396
• 025 - US 2963575
• 026 - US 5677533
• 027 - US 4932784
• 028 - US 5247338
• 029 - US 6288777
• 030 - DE 10150436
• 031 - EP 1186856
• 032 - DE 1267700
• 033 - US 3844513
• 034 - US 4050292
• 035 - US 4,129,276
• 036 - DE 3309908
• 037 - WHERE 8801956
• 038 - US 5133521
• 039 - US 6416020
• 040 - EP 1212228
• 041 - EP 1207091
• 042 - US 5705818
• 043 - US 5636026
• 044 - US 5181327
• 045 - DE 19646098
• 046 - DE 4015086
• 047 - US 5903355
• 048 - US 3206596
• 049 - GB 836721

6.3 named as references Standards, standards, reports and papers

• 050

  UIC Fiche 505-1, "Matériel de transport ferroviaire - Gabarit de construction de materiel roulant "(Rolling Stock Construction Gauge) and current annexes, 9e édition, Août 2002, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  051

  UIC Fiche 505-4, "Consequences of the application of gabarits cinematiques defines par les fiches 505 sur l'implantation des obstacles par rapport aux voies et des voies entre elles "and current annexes, 3e édition, 01.01.1977, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  052

  UIC Fiche 505-5, "Conditions de base communes aux fiches n ° 505-1 à 505-4 - Commentaires sur l'élaboration et la prescriptions de ces fiches "and current annexes, 2e édition, 01.01.1977, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  053

  UIC Leaflet 506, "Rules governing application of the enlarged GA, GB and GC gauges "and current annexes, 1st edition, 01.01.1987, updated on 01.01.1990, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  054

  UIC Leaflet 596-5, "Transport of road vehicles on Wagons - Technical Organization - Method 1 - Transport of ordinary grab-hundled semi-trailers on special wagons (standard recess Wagons) "and current annexes, 3rd edition, 01.07.1985, CODE UIC, UIC, Union International Chemins de fer (International Union of Railways), Paris, France.

  055

  UIC Fiche 596-6, "Trafic de véhicules routiers sur Wagons - Organistion technique - conditions de codification of the unit de chargement en transport combine et the lignes de transport combiné "and current annexes, 3e édition, 01.01.1996, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  056

  UIC Leaflet 597, "Piggyback system - Semi-trailers on bogies - Characteristics ", 1st edition, 01.01.1991" and current annexes, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  057

  UIC Fiche 438-1, "Marquage numérique unifié you matériel remorqué à voyageurs "and current annexes, 2éme édition, 01.01.1988, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  058

  UIC Fiche 438-2, "Marquage numérique unifié you matériel à marchundises "and current annexes, 6e édition, 01.01.1987, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  059

  UIC Fiche 438-3, "Marquage d'identification du matériel moteur "and current annexes, 1ére édition, last amended on 01.01.1971, updated on 01.06.1984, CODE UIC, UIC, Union Internationale des Chemins de fer (International Union of Railways), Paris, France.

  060

  "RIV. Accord sur l'échange et l'utilization wagon entre enterprises ferroviaires ", last amended on 01.10.2001, Union International des Chemins de fer (International Union of Railways), Paris, France.

  061

  "Regulations concerning the International Carriage of Dangerous Goods by Rail (RID) ", 2003 Edition, Intergovernmental Organization for International Carriage by Rail (OTIF), Gryphenhübeliweg 30, CH-3006 Berne, available from TSO Customer Services, PO Box 29, Norwich NR3 1GN, United Kingdom.

  062

  "Inventory of the AGCT and AGC standards and parameters ", United Nations Economic Commission for Europe (UNECE), Working Party on Combined Transport, Geneva, Switzerlund, 1997.

  063

  "Report of the ad hoc multidisciplinary group of experts on the safety in tunnels (rail) on its fourth session (26-27 June, 2003) - Annex", Doc. # TRANS / AC.9 / 8, United Nations, Economic and Social Council, Economic Commission for Europe (UNECE), Inlund Transport Committee, Geneva, Switzerlund, July 30 ^th , 2003.

  064

  "Report of the ad hoc multidisciplinary group of experts on the safety in tunnels (rail) on its first session (27-28 June, 2002) - Addendum 1 - Annex 4 - Report on Safety in Railways Tunnels (as transmitted by the International Union of Railways (UIC)) ", Doc. # TRANS / AC.9 / 2 / Add.1, United Nations, Economic and Social Council, Economic Commission for Europe (UNECE), Inlund Transport Committee, Geneva, Switzerlund, July 24 ^th , 2002.

  065

  A.Lancia, "Infrared scanning Systems for the automatic detection of overheating and incipient fires in trucks approaching major tunnels ", International Conference on Tunnels Safety and ventilation, Graz University of Technology, Austria, 8-10 April 2002 (available as a separate printed article).

  066

  Pigorini B. and Lancia A., "Dispositivi e Sistemi Preventivi per la Riduzione dei Rischi di Incendio nelle Gallerie Stradali e Ferroviarie" (Preventive systems and devices for the reduction of fire-related risks in road and rail tunnels), Acts of the FASTIGI Symposium "Sicurezza delle Gallerie nelle Grundi Infrastructures ed Interoperabilità Europea", Civitavecchia, Italy, July 3 ^rd , 2003.

  067

  Pieralli A., Bracciali A. and Cascini G., "Detettore di Ruota Piatta e Portals in linea per verifica di sagoma ", Acts of the" Convegno SICI ", Napoli, 27-28.11.98, pp. 242-248, SICI (Collegio Ingegneri Ferroviari Italiani), Via Giolitti, 48 - 00185 Roma, Italy, 1998.

  068

  P.W. Kruse and D. D. Skatrud, Editors, "Uncooled Infrared Imaging Arrays and Systems ", Semiconductors and Semimetals, Vol. 47, Academic Press, San Diego, USA, 1997.

  069

  R.A. Wood, T.M. Rezachek, P. W. Kruse and R.N. Schmidt, SPIE Proceedings # 2552, SPIE - The International Society for Optical Engineering, 1000th 20th St., Bellingham WA 98225-6705 USA, 1995

  070

  David W. Embley, "Object Database Development: Concepts and Principles ", Addison-Wesley Pub Co, 1st edition, January 1998.

  071

  François Bancilhon, Claude Delobel, Paris C. Kanellakis (Eds.), "Building an Object-Oriented Database System, The Story of O2 ", Morgan Kaufmann, 1992

  072

  Dennis Fowler, "Virtual Private Networks: Making the Right Connection ", Morgan merchant; 1st edition, June 15, 1999.

  073

  Ruixi Yuan, "Virtual Private Networks: Technologies and Solutions ", Addison-Wesley Pub Co., 1st edition, April 20, 2001.

  074

  David A. Marca, Clement L. McGowan, "City: Structured Analysis and Design Techniques, McGraw-Hill Software Engineering Series, McGraw Hill Text, 1988

  075

  UIC Leaflet 502, "Special consignments - Provisions concerning the preparation and conveyance of special consignments ", 4th edition of 1.7.74 - Reprint dated 1.1.93 and 4 Amendments, CODE UIC, UIC, Union International Chemins de fer (International Union of Railways), Paris, France.

6.4 Reference information about industrial products

• 950

  "RDS80001 / RDS80002 Series High Speed railwheel sensors ", Technical bulletin 100453 EN, Issue 1, Honeywell Automation and Control Products, Newhouse Industrial Estate, Motherwell, Lanarkshire, ML1 5SB, Scotlund, United Kingdom, 2003.

  951

  "Introduction to Variable Reluctance sensor (VRS)," Invensys Sensor Systems / Electro Corporation (currently a part of Honeywell), 1845 57 ^th St., Sarasota, FL, USA., 2003

  952

  "Optocator Laser Sensors, Product Information", LMI Selcom AB, Box 250, SE-433 25 Partille, Sweden, 2003

  953

  "Optocator Interface Module. Product Information ", LMI Selcom AB, Box 250, SE-433 25 Partille, Sweden, 2003

  954

  "Laser Sensors for Road", LMI Selcom AB, Box 250, SE-433 25 Partille, Sweden, September 29 ^th , 2003

  955

  "Line Scan Cameras Piranha CL-P1", Product Datasheet, DALSA Corporation, 605 McMurray Road, Waterloo, Ontario, Canada N2V 2E9, 2003

  956

  "Advanced Line Scan Cameras Eclipse EC-11", Product Datasheet, DALSA Corporation, 605 McMurray Road, Waterloo, Ontario, Canada N2V 2E9, 2003

  957

  "DALSA HS-41 Advanced Line Scan Cameras", Product Datasheet, DALSA Corporation, 605 McMurray Road, Waterloo, Ontario, Canada N2V 2E9, 2004

  958

  "Specifications of the LD90-3100VHS-FLP", Data Sheet, Riegl Laser Measurement Systems GmbH, Riedenburgstrasse 48, A-3580 Horn, Austria, September 2001

  959

  "Specifications of the LD90-3100EHS-FLP", Data Sheet, Riegl Laser Measurement Systems GmbH, Riedenburgstrasse 48, A-3580 Horn, Austria, September 2001

  960

  "Short-Range Airborne Laser Scanner LMS-Q140i-60/80", Data sheet, Riegl Laser Measurement Systems GmbH, Riedenburgstrasse 48, A-3580 Horn, Austria, October 2002

  961

  "Technical data PROFILER / ILAR", Technical data sheet, Zoller + Fröhlich GmbH Electrical Engineering, Simoniusstraße, 22, Wangen im Allgäu, Germany, 2000th

  962

  "Sentry System", Hot Box Detector, Product information, Southern Technologies Corporation, 6145 Preservation Drive, Chattanooga, TN 37416, USA, 2003.

  963

  Products information web pages at www.getransportation.com, General Electric Transportation Systems, Erie, PA, USA, December, 2003.

  964

  "Hot Box and Brake Detection System VAE-HOA / FO A400 ", Product Bulletin, VAE Eisenbahnsysteme GmbH, A-8740 Zeltweg, Alpinestraße 1, Austria, Of 2003.

  965

  "M-2105 Series PbSe Focal Plane Arrays", Product Leaflet, Northrop Grumman Corporation, Electro-Optical Systems, 1215 p. 52 ^nd Street, Tempe, AZ, USA, 85281, 2003.

  966

  "MULTIRAIL ^® WheelScan" Product information bulletin BV-D 5004 GB, SCHENCK PROCESS GmbH, Lund Wehrstraße, 55, D-64273, Darmstadt, Germany, of 2003.

  967

  WheelChex ^™ , Product information bulletin AEATRD / 803 (9/00), AEA Technology Rail, PO Box 2, rtc Business Park, London Road, Derby DE24 8YB, United Kingdom, 2000.

  968

  PANCHEX ^® , Product Information http://www.aeat.co.uk, AEA Technology Rail, PO Box 2, rtc Business Park, London Road, Derby DE24 8YB, United Kingdom, 2003

  969

  NI PCI-7041 / 6040E Real-Time Multifunction Data Acquisition Board ", Technical Bulletin, National Instrument Corporation, 11500 N Mopac Expwy, Austin, TX 78759-3504, USA, 2003

  970

  "PC-DIO-96", Technical Bulletin, National Instruments Corporation, 11500N Mopac Expwy, Austin, TX 78759-3504, USA, 2003

  971

  "VMEbus products guide", Motorola computer Group, 2900 S. Diablo Way, Tempe, AZ, 85282, USA, 2003

  972

  "CompactPCI Catalog", Gespac SA, 18 Chemin des Aulx, 1228 Geneva, Switzerlund, 2003

  973

  Cisco VPN 3000 Concentrator, Data Sheet, Cisco Systems, Inc., 170 West Tasman Drive, San Jose, CA 95134-1706, USA, 2004

  974

  "Cisco VPN 3002 Hardware Client", Data Sheet, Cisco Systems, Inc., 170 West Tasman Drive, San Jose, CA 95134-1706, USA, 2001

  975

  ORBCOMM, 21700 Atlantic Boulevard, Dulles, VA 20166, USA

  976

  "KX-G7100 ORBCOMM subscriber communicator", Product Specification Datasheet, Matsushita Electric Corporation of America, One Panasonic Way, Secaucus, NJ 07094, USA, 2003

  977

  "eMGW Wireless DSL for Data and Telephony Services, "Alvarion Ltd., International Corporate Headquarters, 21a HaBarzel Street, P.O. Box 13139, Tel Aviv, 61131, Israel, 2003

  978

  "SA-PCD BreezeNET Pro.11 PCMCIA Cards", Technical Datasheet, Alvarion Ltd., International Corporate Headquarters, 21a HaBarzel Street, P.O. Box 13139, Tel Aviv, 61131, Israel, 2003

  979

  "RTLinuxPRO 2.0", FSMLabs, Inc., 115 D Abeyta Ave, Socorro, NM 87801, USA, 2003

6.5 Assignment of reference numbers by areas

• 001-048: Reference numbers of cited patent documents
• 050 to 075: Reference numbers of public Standards and papers
• 100-145: Reference numbers of the mathematical formulas
• 151 - 167 : Reference numbers in 1
• 201 - 217 : Reference numbers in 14
• 218 - 228 : Reference numbers in 4
• 230 - 246 : Reference numbers in 3
• 250 - 267 : Reference numbers in 2
• 275 - 301 : Reference numbers in 5
• 310 - 325 : Reference numbers in 6
• 345 - 364 : Reference numbers in 8th
• 369 - 399 : Reference numbers in 9
• 400 - 430 : Reference numbers in 7
• 440 - 471 : Reference numbers in 10
• 480 - 505 : Reference numbers in 11
• 520 - 531 : Reference numbers in 12
• 540 - 551 : Reference numbers in 13
• 560 - 585 : Reference numbers in 15
• 600 - 625 : Reference numbers in 16
• 630 - 660 : Reference numbers in 17
• 670 - 697 : Reference numbers in 18
• 700 - 730 : Reference numbers in 19
• 740 - 752 : Reference numbers in 20
• 760 - 777 : Reference numbers in 21
• 780 - 807 : Reference numbers in 22
• 810 - 853 : Reference numbers in 23
• 860 - 885 : Reference numbers in 24
• 950 to 978: Reference numbers of information sheets and notices in terms of commercially distributed products

Claims (58)

Method for detecting and signaling at least one faulty and / or dangerous condition for train formation ( 151 ) of at least one passing rail vehicle ( 152 ), the at least one faulty or dangerous condition in particular comprising: danger of the gauge line, shifted loads, overheating, failures and incipient failures in the axle bearings, overheating of the wheels and brakes, overheating of parts of the vehicle body and fire on board, each passing Rail vehicle is a corresponding execution model of a variety of predetermined execution models; the method performing at least the following operations: a. Data transfer ( 154 ) of sensors and instruments ( 153 ) positioned along and / or around a track section and electronically storing a data set ( 156 ), which carry out a large number of measurements with respect to a passing rail vehicle ( 155 ) such that a time value can be directly or indirectly associated with each measurement from the plurality of measurements, the passing rail vehicle continuing on the rail section; b. Determining the execution model ( 158 ) of the passing rail vehicle through the processing ( 157 ) at least part of the data amount and data characteristics of the plurality of vehicle execution models; c. Recall ( 159 ) of vehicle-specific information and data ( 162 ) stored in at least one database ( 161 ) are stored and linked to the determined execution model; d. Processing ( 160 ) of at least part of the vehicle-specific information and data for calculating a parameter set ( 163 ) defining one or more mathematical functions as a function of the time, position and orientation of at least one basic component of the passing rail vehicle to associate at least a portion of the amount of data with corresponding portions of the passing rail vehicle or its cargo; e. To capture ( 164 ) of at least one erroneous and / or dangerous condition for the passing rail vehicle based on the amount of data using mathematical functions and at least part of the vehicle-specific information and data; f. Produce ( 166 ) of alarm signals ( 167 ) for a rail signaling and safety system to prevent accidents and to reduce losses and costs if at least one faulty and / or dangerous condition ( 165 ) is detected. Method according to claim 1, where the operation b contains the step, distances between the Wheel groups of train formation together with data and information, the vehicle design models match, for the decision to use one or more execution models correspond to the passing rail vehicle. Method according to claim 2, the step of compiling a list of vehicle execution models comprehensive, which may correspond to the passing rail vehicle. Method according to claim 1, where operation b is the application of one or more vehicle information algorithms including the optical character recognition of at least a part of the symbols includes within the standard tag codes on the pages the rail vehicles are mounted. Method according to claim 4, wherein the optical character recognition for at least one on the side of passing rail markings and a mark carried out which is on the other side of the particular passing rail vehicle is appropriate. The method of claim 1, wherein the operation b includes the step of data corresponding to the weight or the load on the rails of wheels or wheel groups of the passing rail vehicle to use for deciding whether one or more execution models correspond to the passing rail vehicle. Method according to claim 1, wherein the operation b comprises the step, the diameters of the wheels and / or of one or more profiles ( 362 . 363 ) on the outer surfaces of the wheels of the passing rail vehicle for deciding whether one or more execution models correspond to the passing rail vehicle, such as profiles and wheel diameters resulting from the processing of the data acquired in operation a from electro-optical sensors and instruments were. Method according to claim 1, where the operation b includes the step, one or more surface profiles along the length train train to use for the decision whether one or more of the execution models correspond to the passing rail vehicle, such as profiles, the arising from the processing of data obtained in operation a were taken over by electro-optical sensors and instruments. Method according to claim 8, where the surface profiles determined by measurements made with laser distance meters, those at a height and are mounted in an alignment and so the profiles of Include buffers of rail vehicles. Method according to claim 1, wherein the operation b comprises the step, one or more visual or geometric characteristics of a vehicle design model for the decision to use one or more of the execution models to pass Rail vehicle by processing the data, which were taken over in operation a by electro-optical sensors and instruments. Method according to claim 1, the calculation of distances between the wheel groups and one comprising mathematical function representing the longitudinal position of the passing Rail vehicle along the track over time through a Algorithm that uses data in operation a for a series of Wheel sensors adopted were. The method of claim 11, wherein the calculation algorithm is based on the data acquired for a series of wheel sensor pairs, wherein two sensors ( 312 . 313 ) of each of these pairs are mounted on a different rail and are aligned substantially along the axis perpendicular to the rails. Method according to claim 12, where the distances between the wheel groups and the mathematical Function the longitudinal position of the passing rail vehicle along the track over the Time represent and by the positions of the centers of the wheel groups are defined over time. Method according to claim 13, where the distances between the wheel groups and the mathematical Feature the longitudinal position representing the passing rail vehicle along the track over time, be calculated by minimizing a mathematical function whose expression is double summations over a number of K values a longitudinal distance from wheel sensors or wheel sensor pairs along the rail of one Origin and over a number of J values of the longitudinal distance the wheel center of a certain origin on the contains passing rail vehicle, each of the expressions of this double summations a monotonous non-negative function of algebraic sum with the corresponding sign of at least the longitudinal distance a wheel sensor or wheel sensor pair value, a longitudinal distance a wheelgroup midpoint value and the value corresponding to one Time value of the mathematical function corresponds to the longitudinal position of the passing rail vehicle along the track over time represents. Method according to claim 14, the mathematical function being the longitudinal position of the passing Rail vehicle along the track over time, one piece at a time Function is, with continuity conditions the first and second derivative, and in particular a spline function, which consists of cubic polynomials. Process according to claims 14 or 15, wherein the minimization under the appropriate mathematical constraints carried out will that for the maximum values of the acceleration and / or deceleration of the passing Rail vehicle apply. Method according to claim 1, where operation b detecting the unique identity of the passing Rail vehicle includes. Method according to claim 17, the determination of the unique identity of a Vehicle on the optical character recognition of at least one Part of the icons based within the standard marker codes attached to the sides of the rail vehicle. Method according to claim 1, where operation b is performed is by adding more information about the passing rail vehicle considered Be the other information by reading or querying the devices, which are attached to the passing rail vehicle, or of one or more rail information systems. The method of claim 1, wherein operation d ( 160 ) a step for defining the position and orientation of a body ( 250 ) of the passing rail vehicle over time by means of a coordinate transformation function, wherein the coordinate transformation function also the definition in a first coordinate system ( 257 . 260 . 255 . 258 integral with the vehicle body of a vector / point ( 263 ), which in a second coordinate system ( 253 . 256 . 259 . 254 ) integral with sensors and instruments ( 153 ) is defined along the path, wherein the coordinate transformation function is uniquely associated with an inverse coordinate transformation function that allows the inverse coordinate transformation of a vector / point. Method according to claim 20, where for the calculation of the coordinate transformation function the mathematical Function is used, which is the longitudinal position of the passing Rail vehicle along a track over time represents. A method according to claim 20 or 21, wherein the coordinate transformation function is defined by a set of parameters ( 163 ) obtained by minimizing a function ( 160 ) whose mathematical expression comprises the summation of a series of non-negative expressions, each term representing the correspondence between one or more data from the set of inherited data with a geometric feature of the vehicle body ( 250 ) according to the relevant vehicle-specific information and data from the vehicle database ( 161 ). The method of claim 22, wherein a non-negative term or terms are a function of the distance between a vector / point in a three-dimensional space, such as a vector / point obtained from at least one measure of the data set, and a Area in the same three-dimensional space, where the area corresponds to a physical characteristic of the passing rail vehicle, and in which vehicle-specific information and data are defined, and such a three-dimensional space either integral with the vehicle body or the sensors and instruments ( 153 ) is along the way. A method according to claims 20, 21, 22 or 23, wherein the operation e ( 164 ) the detection step of limit line related hazards and / or hazards due to displaced loads ( 237 ) for the passing rail vehicle and its cargo by comparing the coordinates of vectors / points ( 616 ) derived from the data acquired by the sensors and instruments ( 153 . 460 and or 497 and or 528 and or 547 ) with one or more vehicle-specific surface (s) or profile group (s) ( 608 - 612 ), which are in a coordinate system ( 604 . 605 . 606 . 607 ) integral with the vehicle body ( 250 ), where the three-dimensional surfaces or profile groups are defined either by: - information and data ( 162 ) from the vehicle database ( 161 ), or - calculation based on the information and data from the vehicle database, or - information and data from the vehicle database ( 161 ) together with the data and information corresponding to those obtained from the recognition of particular markings on the vehicle and / or cargo based on the data taken from the sensors and instruments, and the comparison is performed in three - dimensional space, is integral with the vehicle body or with the sensors and instruments along the path, by using the coordinate transformation functions or their inverse functions. A method according to claim 24, wherein an alarm ( 167 ) to the rail safety and signaling system ( 845 ) is sent ( 166 ), if the comparison of claim 24 detects a hazard associated with one or more vectors / points ( 616 ), which is outside the vehicle speci three-dimensional bounding surfaces or profile groups ( 562 . 582 . 583 . 577 ) defined in accordance with the principles of the UIC Code, and in particular with the Leaflets 505-1, 505-4 and 505-5, which contain a reference boundary line profile ( 563 ) and / or a profile for concrete obstacles ( 573 ) of a track section and also take into account the measurement accuracy and tolerances defined by the relevant railway company. Method according to claim 25, where the information about a planned speed for the train formation for the definition of the comparison tolerances and / or the definition of vehicle-specific three-dimensional surfaces or Profile groups are used, with information about the planned speed come from a rail information system. A method according to claim 24, wherein an alarm signal ( 167 ) to the rail safety and signaling system ( 845 ) is sent ( 166 ), if the comparison of claim 24 detects a hazard associated with one or more vectors / points ( 616 ), which are outside of the vehicle-specific three-dimensional delimiting areas or profile groups corresponding to a cargo profile consistent with markings on the vehicle and / or the loads, in particular according to the relevant specifications of UIC codes 502, 595-5, 595 -6 and 597, which take into account the accuracy of the measurements and calculations as well as the tolerances defined by the relevant railway company. Method according to claim 24, wherein the three-dimensional areas or profile groups are provided with freight profiles of the vehicle, in accordance with certain applicable rules and regulations, such as freight profiles and corresponding rules preferably defined in the RIV agreement, or with an execution profile ( 585 ) of the vehicle. The method according to claim 20, 21, 22 or 23, wherein the operation e ( 164 ) includes detecting the violation of a cargo loading rule, such as preferably a cargo loading rule defined in the RIV Agreement, for a passing rail vehicle and one or more adjacent vehicles with respect to the position of one or more loads by the processing of coordinates of the vectors / points ( 616 ) obtained from the data acquired by the sensors and instruments and using the coordinate transformation functions or their inverse functions and vehicle-specific information and data from the vehicle database. Method according to claim 24, 28 or 29, with a warning or signaling on a track safety system is issued, along with data and information with which an operator or a software application the questionable or possible Danger can better evaluate, if a danger or a possible danger is detected. A method according to claim 24, wherein the detection of limit line related hazards and / or hazards due to shifted cargo ( 237 ) for a passing rail vehicle and the freight comprises the steps of the coordinates of vectors / points ( 616 ) of at least two datasets generated by sensors and instruments ( 153 . 460 and or 497 and or 528 and or 547 ), wherein each pair of at least two sets of data corresponding to groups of positions longitudinally spaced on the vehicle by at least one meter and requiring the same risk to be consistently detected on the basis of all of at least two sets of data. The method of claim 1, wherein the operation d ( 160 ) a definition step of the position and orientation of an axis ( 683 . 696 ) of the passing rail vehicle and / or the central axis of the axle and / or of components such as wheels ( 644 . 631 ), Brakes ( 643 . 656 . 634 . 632 . 674 . 690 ) and bearings ( 654 . 630 ) associated with the axis by an axis-related coordinate transformation function over time, this axis-related coordinate transformation function defining the definition in a coordinate system ( 679 . 680 . 675 . 678 . 685 . 688 . 689 ) integral with the components of such an axis ( 683 . 696 ) of a vector / point ( 670 . 692 . 637 . 639 . 640 . 641 . 646 ), which are in a coordinate system ( 682 . 681 . 677 . 697 . 687 . 684 ) are defined integrally with the sensors and instruments along the path, the axis-related coordinate transformation function being uniquely associated with an inverse axis-related coordinate transformation function that permits the inverse coordinate transformation of a vector / point. The method of claim 32, wherein the axis-related coordinate transformation function is defined by a set of parameters obtained by minimizing a function whose mathematical value a term comprising the summation of a series of non-negative terms, each of which expresses the correspondence between one or more data of the amount of inherited data and the position of the corresponding axis-related components according to the corresponding vehicle-specific information and data from the vehicle database ( 161 ). Method according to claim 32 or 33, wherein the calculation of the axis-related coordinate transformation function the mathematical function uses the longitudinal position of the corresponding passing Rail vehicle along the track over time as an indication the longitudinal position and / or Speed and / or acceleration of the corresponding wheel center points over time. Method according to claims 32, 33 or 34, wherein the operation e ( 164 ) the detection step of erroneous and / or dangerous conditions for ( 239 ) contains the bearings and / or brakes and / or wheels of the axle in question, through the use of algorithms for processing part of the data volume generated by sensors and instruments ( 650 . 638 ), which sense the infrared radiation radiated from the bearings and / or brakes and / or wheels of the relevant axis, these algorithms representing the axis-related coordinate transformation function and / or the inverse function thereof together with information and data from the vehicle database ( 161 ) to associate the respective inherited data with elements of the corresponding axle and / or bearings and / or wheels and / or brakes. Method according to claim 35, bearing failure and / or abnormal heating of the Wheels and / or be detected by parts of the brake, with statistical significance criteria be applied to the data, based on the element temperatures of bearings and / or wheels and / or brakes, wherein the elements of a stock of Brakes and / or wheels and / or brakes corresponding to the passing rail vehicle or other vehicles in the train formation, the same execution model such as the passing rail vehicle, and possibly the application the statistical significance criteria on a subset of the population restrict, according to the components, on only one side of the train formation are attached. Method according to claim 36, where the temperature-related data for certain components of Bearings are calculated by an algorithm that measures the weight and / or the freight data, the for the corresponding wheels or wheel groups were taken into account, and / or the data the for the temperature of the rail network and / or the wheel hub of the relevant Wheels representative are. A method according to claims 20, 21, 22 or 23, wherein the operation e ( 164 ) capturing ( 238 ) on board, as well as thermally detectable hazards to the passing rail vehicle body and its cargo, through the use of algorithms to process part of the volume of data acquired from sensors and instruments, and in particular from those sensors and instruments ( 760 . 768 ) that sense the infrared radiation emitted by the vehicle body and / or interior elements of the vehicle body and / or cargo of the passing rail vehicle, these algorithms representing the coordinate transformation functions for components of the vehicle body and / or their inverse function together with information and data the vehicle database ( 161 ) to associate the respective inherited data with components of the vehicle body and / or the interior elements of the vehicle body and / or cargo of the passing rail vehicle. A method according to claim 38, wherein the vehicle-specific information and data from the vehicle database ( 161 ) one or more algorithms of a set of changing algorithms for detecting ( 238 ) define fire on board and thermally detectable hazards to the body of the passing railcar and its cargo, and also define the parameters for the application of these algorithms, including in particular the parameters defining the surface elements associated with the vehicle and its cargo a coordinate system that is integral with the vehicle body. Method according to claim 1, wherein at least the operations a, b, c, d, e and f for all vehicles in the train formation ( 151 ) except for those vehicles for which vehicle-specific information and data from the vehicle operation indicate not to perform certain operations or parts of certain operations and / or for those vehicles for which the corresponding execution model can not be determined. A method according to claim 40, wherein the detection of certain defects and hazardous conditions for those vehicles for which the corresponding execution model can not be determined is performed by methods that do not include vehicle-specific information and data from the vehicle database ( 161 ) use. System for carrying out the method of detecting and signaling at least one faulty and / or dangerous condition for train formation ( 151 . 204 ) of at least one of claims 1 to 41, the system comprising: - sensors and instruments ( 153 . 884 . 881 . 875 . 869 . 860 . 877 . 871 . 872 . 862 . 873 . 870 . 876 . 614 . 784 . 792 . 800 ) along and / or around a track section ( 201 . 202 . 203 ) for generating measured data ( 156 ), the elements of rail vehicles ( 204 ), which during the movement of the train formation in this section has a spatial area ( 211 to cross; - first hardware ( 780 . 786 . 789 . 799 . 811 ) and software device ( 226 . 218 . 233 . 232 ) for takeover ( 154 ) of a dataset ( 156 ) of the sensors and instruments and of the electronic storage, which are the subject of a large number of measurements relating to a passing rail vehicle ( 155 ) in such a way that a time value ( 781 ) can be linked, directly or indirectly, to any measurement from that plurality, such hardware and software equipment being adapted to receive data from other sensors, instruments and systems; Second hardware ( 810 - 813 . 780 . 786 . 789 . 799 ) and software device ( 234 - 239 . 246 ) for the electronic processing of data ( 156 ) taken from the sensors and instruments and other data and information; - third hardware ( 821 . 817 . 823 . 853 ) and software device for connecting and generating alarms for rail signaling and security systems ( 845 . 843 . 844 ); Connection hardware ( 814 . 795 . 796 . 797 . 798 . 807 . 782 . 783 . 787 . 793 . 805 . 806 . 785 . 790 . 791 . 801 . 802 . 803 . 213 ) and software means for connecting and powering the sensors and instruments and the first, second and third means; - Means for positioning and shielding the sensors and instruments and the first, second and third and the connection hardware. A system according to claim 42, wherein further sensors and instruments, in particular wheel sensors, connected to the first device are located at one or two positions ( 206 . 207 ) at a distance ( 210 . 212 ) of the spatial area ( 211 ) to detect the arrival of the train formation and to prepare the system for operation a. A system according to claims 42 or 43, wherein the sensors and instruments comprise a pair of wheel sensors ( 312 . 313 ) which are installed to detect the passage of both wheels in a wheel group at a substantially same longitudinal position along the track. A system according to claims 42 or 43, wherein the sensors and instruments comprise one or more laser distance measurement sensors ( 350 . 873 ) for obtaining profiles ( 358 . 362 . 363 ) of the wheels ( 345 . 346 ) and / or parts of the lower part of a passing rail vehicle, in particular the buffer ( 295 ). A system according to claims 42 or 43, wherein the sensors and instruments comprise linearly arranged recording devices ( 440 to 447 . 449 . 450 . 460 . 468 . 770 . 870 . 876 ) to capture rows of shooting data corresponding to the surface elements of a passing rail vehicle, such elements being approximately in a plane that is substantially vertical at the time of shooting, with corresponding artificial lighting devices (FIGS. 461 . 463 ) to be provided. System according to claims 42 or 43, wherein the sensors and instruments are laser distance measuring units and in particular groups of fixed laser distance measuring devices ( 481 to 492 . 493 . 494 . 497 . 498 . 499 ) and / or time-of-flight or triangulation-type laser distance measuring devices ( 520 . 525 . 526 . 527 . 528 ) and / or distance laser measuring devices with fast sampling based on the laser modulation phase detection ( 540 . 547 . 766 . 769 . 860 . 869 . 875 . 881 ) to detect series of measurements of the three-dimensional position of surface elements of a passing rail vehicle and its cargo. A system according to claim 42 or 43, wherein the sensors and instruments comprise means for sensing the heat radiation emitted by the axle-related components, and more particularly from the bearing housings (10). 654 . 630 ), Wheels ( 644 . 631 ) and brake parts ( 643 . 656 . 634 . 632 ), Where in each of these heat radiation sensing devices, comprising a single infrared radiation detecting device, in particular with an opto-mechanical scanning system or infrared radiation or linear or multilinear series sensing elements ( 650 . 638 . 862 . 872 ) scanning elements of the infrared radiation, such as rows of uncooled micro-thermopile sensors or thermoelectrically cooled lead-selenide photoconductor sensors or cooled mercury and cadmium telluride or indium antimonide photon sensors. A system according to claim 42 or 43, wherein the sensors and instruments comprise means for sensing the thermal radiation radiated from the body of a passing railcar and / or cargo, each of said heat radiation sensing means comprising at least one photon detector together with an opto-mechanical scanning system or a linear, multilinear or two-dimensional series ( 650 . 638 . 862 . 872 ) of infrared radiation sensing elements ( 871 . 877 ), such as rows of uncooled micro-thermopile sensors or thermoelectrically-cooled lead-selenide photoconductor sensors or cooled mercury and cadmium telluride or indium antimonide photon sensors. A system according to claim 42 or 43, wherein a system for measuring ( 867 . 864 ) the weight or load on the track for the wheels or wheel groups of a passing rail vehicle is connected to the first or second device using the weight or cargo measurements according to the method of claim 37 and / or to generate alarms or alerts; if one or more of these weights or loads are dangerous or undesirable according to the vehicle-specific information from the vehicle database ( 161 ) and / or according to one or more parameters for the relevant track section. A system according to claim 42 or 43, wherein the measurement rate and / or the data acquisition rate for one or more sensors and instruments is set and / or changed ( 780 . 218 . 232 ) according to the speed of train formation of the rail vehicle, such speed being evaluated by wheel sensor signals or other suitable means. A system according to claim 42 or 43, wherein at least a part of the first device ( 780 . 786 . 789 . 799 ) and the second facility ( 810 to 813 . 821 ) by the connection device ( 795 . 796 . 797 . 807 . 814 ) in the form of a local ( 815 ) Network ( 798 ) are connected. A system according to claim 42 or 43, wherein at least a part of the second device ( 810 to 813 . 821 ) directly or indirectly by cable and / or optical fiber and / or radio links and / or in particular through the Internet ( 824 ) with remote data processing centers ( 834 ) and / or rail safety and signaling data processing systems ( 843 ) and / or with other rail data processing systems ( 852 ) connected is. System according to claim 53 and 46, a software application comprising the image data the stop notice for Hazardous goods and by optical character recognition the corresponding Codes recognizes and the dangerous goods information with the appropriate Vehicle and, if applicable, with the net weight of the dangerous goods, the are loaded on the passing rail vehicle, linked this information for one or more vehicles in a train a lot of information form, sent to other systems or for possible later access by other systems stored or used to generate alarms for the rail safety and signaling system can be used if the relevant Transport of dangerous goods not one or more restrictions for the Traffic on the section of track concerned. A system according to claim 53, wherein one with a radio communication device ( 847 ) portable data processing unit ( 846 ) Data and information from a system installation ( 815 ) to assist those responsible for identifying a vehicle in which this system installation has detected one or more faulty and / or dangerous conditions and in carrying out the actions concerned that must be performed by that personnel. The system of claim 53, wherein a remote system management center or multiple remote system management centers ( 834 ) with a system installation ( 815 ) maintain the software and the data volumes of such a system installation and / or monitor the correct functioning of such a system installation, whereby satellite ( 825 ) and terrestrial ( 826 ) Communication systems can be used. The system according to claim 42 of 43 and the method of claim 17, wherein the information about erroneous or dangerous conditions detected for a passing rail vehicle, and in particular for those defects that do not correspond to a serious immediate danger, with the unique identity of the vehicle linked rail vehicle and directly or indirectly ( 834 . 842 ) the information systems related to the maintenance of wagon and locomotives ( 842 ) to provide. A system according to claim 42 or 43 and the method of claim 17, wherein the unique identity of the passing rail vehicle is directly or indirectly ( 834 . 842 ) Information systems ( 842 ) of the rail transport logistics is made available.