WO2018087341A1 - Procédé d'inspection, système de traitement de données et système d'inspection servant à inspecter un véhicule à l'état de fonctionnement - Google Patents

Procédé d'inspection, système de traitement de données et système d'inspection servant à inspecter un véhicule à l'état de fonctionnement Download PDF

Info

Publication number
WO2018087341A1
WO2018087341A1 PCT/EP2017/078967 EP2017078967W WO2018087341A1 WO 2018087341 A1 WO2018087341 A1 WO 2018087341A1 EP 2017078967 W EP2017078967 W EP 2017078967W WO 2018087341 A1 WO2018087341 A1 WO 2018087341A1
Authority
WO
WIPO (PCT)
Prior art keywords
inspection
vehicle
data
sensor
inspection system
Prior art date
Application number
PCT/EP2017/078967
Other languages
German (de)
English (en)
Inventor
Ronny SÖLLNER
Karl-Heinz Förderer
Stefan SCHÖLLMANN
Original Assignee
Deutsche Bahn Fernverkehr Ag
Psi Technics Gmbh
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Deutsche Bahn Fernverkehr Ag, Psi Technics Gmbh filed Critical Deutsche Bahn Fernverkehr Ag
Publication of WO2018087341A1 publication Critical patent/WO2018087341A1/fr
Priority to PCT/EP2018/080950 priority Critical patent/WO2019092248A1/fr

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B61—RAILWAYS
    • B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
    • B61L27/00—Central railway traffic control systems; Trackside control; Communication systems specially adapted therefor
    • B61L27/30—Trackside multiple control systems, e.g. switch-over between different systems
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B61—RAILWAYS
    • B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
    • B61L27/00—Central railway traffic control systems; Trackside control; Communication systems specially adapted therefor
    • B61L27/10—Operations, e.g. scheduling or time tables
    • B61L27/16—Trackside optimisation of vehicle or train operation
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B61—RAILWAYS
    • B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
    • B61L27/00—Central railway traffic control systems; Trackside control; Communication systems specially adapted therefor
    • B61L27/40—Handling position reports or trackside vehicle data
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B61—RAILWAYS
    • B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
    • B61L27/00—Central railway traffic control systems; Trackside control; Communication systems specially adapted therefor
    • B61L27/50—Trackside diagnosis or maintenance, e.g. software upgrades
    • B61L27/57—Trackside diagnosis or maintenance, e.g. software upgrades for vehicles or trains, e.g. trackside supervision of train conditions

Definitions

  • the invention relates to an inspection method for optically inspecting a vehicle in the operating state using an inspection system comprising a sensor arrangement with one or more sensors, one or more of which are optical sensors, and providing a passage for the vehicle during a relative movement between the vehicle and the sensor assembly.
  • the inspection method includes the step of applying the sensor assembly to at least one inspection portion of the vehicle to inspect the vehicle.
  • the invention further relates to a data processing system which preferably comprises both an inspection system internal and an external, remotely located subsystem.
  • the invention also relates to the inspection system. From the prior art, the aforementioned method, the aforementioned data processing system and the aforementioned inspection system are known, for example, toll bridges or video-based traffic monitoring.
  • the vehicle approaches the toll bridge and the sensor assembly, which serves as optical sensors
  • Data processing system and inspection method can be achieved using at least one optical sensor.
  • the object of the present invention provides an inspection method for inspecting a vehicle in the operating state which accomplishes this object according to the invention
  • the object is achieved by a data processing system having the features of claim 14.
  • the object is achieved by an inspection system having the features of claim 15.
  • the inventors have recognized that the versatility of an inspection system, the
  • Data processing system and the inspection method can be drastically increased if at least one optical sensor of the sensor assembly is used to inspect a technical operating state of the vehicle.
  • Rail vehicles an inspection in operation can bring great benefits, since a train fleet can be inspected in regular driving without being taken to the inspection for the operation.
  • subsystem is understood to mean that the subsystem forms part of the inspection system.
  • the vehicle is a
  • An achievable advantage of the invention is the provision of time-efficient and cost-effective organizational structures for inspection and maintenance of vehicles, such as e.g. Trains, and railway networks. Particularly preferred is a
  • High voltage power line is supplied, such as a passenger train, a freight train, a tram, a subway or a rapid-transit railway.
  • High voltage power supply is often very personal and time consuming. If it is necessary to inspect elements associated with the high voltage, it has hitherto often been necessary to use people who may be exposed to a health hazard as a result of the high voltage. In addition, for the examination of high-voltage associated components, such vehicles often need to be removed for extended periods of inspection
  • An alternative vehicle is a watercraft, such as a ship or a submarine, or an aircraft, such as a helicopter or an airplane.
  • the vehicle is a passage passing vehicle.
  • the vehicle is additionally or alternatively inspected before or after passing through the inspection system in the operating state.
  • an inspection section has an inspection object, also called an inspection element or inspection object.
  • the inspection object is a specific component to be inspected in the inspection section.
  • the Inspection section may also have several individual inspection objects to be inspected together.
  • the inspection section may thus be the entire vehicle or may include only a portion of the vehicle or an item of the vehicle.
  • An inspection section is preferably an element selected from the group consisting of so-called roof garden of the vehicle, underbody of the vehicle, air conditioning of the vehicle, individual car of the vehicle, several cars of the vehicle, pantograph of the vehicle, sliding strip of the vehicle
  • the inspection method provides for providing at least one
  • the sensor is preferably selected from the group consisting of optical sensor, olfactory sensor, chemical sensor, vibration sensor , electromagnetic sensor, thermal image sensor and acoustic sensor, wherein at least one provided sensor is an optical sensor, which is preferably triggered, and preferably another sensor is provided in the sensor arrangement, preferably in a measurement technique used by him for collecting raw data from the optical sensor is different.
  • a camera as an optical sensor.
  • a preferred camera is a line camera. In front of the camera, a prism for spectral separation may be upstream.
  • the spectral range of the camera is preferably the range of visible light and the IR range.
  • the camera has a preferred operating temperature range of -20 ° C to 60 ° C for outdoor use.
  • a preferred stand-by temperature is set at about 50-60 ° C, while a preferred detection temperature is between 20-30 ° C.
  • a focal length range may be determined based on field of view, inspection object size, distance, and sensor size.
  • a camera heater can be provided for improved operational readiness and against image noise. Against Störlichtarion can in
  • Embodiments diaphragm elements may be provided.
  • for the Camera provided active camera cooling or passive camera cooling to the inspection system.
  • the passive camera cooling is preferably on
  • the active camera cooling preferably comprises a Peltier element, a camera control or a solar cell. Active camera cooling may also include a fan or water cooling. A preferred minimum resolution of the camera is at least 4 mm. Preferably, the camera is adapted to detect fluids and liquids, in particular at least one of the group consisting of gas, steam, oil, lubricant, condensed water and refrigerant. 3.1.3 Anti-fog device
  • optical elements can be any optical elements in the optical path between an imaging sensor and the object to be inspected.
  • optical elements can be any optical elements.
  • Camera lenses transparent covers, mirrors, prisms, screens or optical filters.
  • Possible devices for anti-fog are, for example, a
  • Non-stick coating for example with lotus effect
  • a spray nozzle for deflecting the optical element with a cleaning liquid and / or a windshield wiper for example, a blower, a heating of the housing of an imaging sensor and / or a
  • Cover for example, a controllable shutter or a preferably electrically controllable iris.
  • an optical deflection element can be mounted in the optical path of the camera, which is provided with a movement device according to the invention.
  • a movement device for example, if a pivot axis and an axis of rotation extend through the optical path of the camera, the optical path can be achieved by pivoting or rotating the optical element be aligned with the inspected section of the moving object.
  • the optical deflection element may be, for example, a mirror.
  • a plurality of cameras can be directed at different angles to an optical deflection element, such as a mirror.
  • an optical deflection element such as a mirror.
  • each camera takes pictures from different perspectives, which can be evaluated by means of digital image analysis into a three-dimensional image of an object of inspection.
  • optical deflection elements such as mirrors, preferably concave or convex curved mirror on train, for example on
  • Bogie itself, to be attached.
  • the straight mirrors are particularly favorable and can easily be attached to a train
  • Implementation system include in particular:
  • Mirror in particular with concave or convex shape with respect to at least one spatial direction, in order to prepare the optical field of view or for focusing. If such a form of the mirror is known, it can be used in an algorithm for
  • Polfilter are adjusted so that due to its selected polarization substantially only by inspecting the portion of the moving object reflected light is completely meditated by the polarizing filter to the optical
  • An optical depth of field adjustment device such as an adjustable iris such as an iris to adjust the optical aperture.
  • a shallower depth of field is beneficial for obtaining information about the spatial resolution of an object.
  • a higher depth of field is advantageous in order to be able to image a section of an object with a varying height profile in an image as sharply as possible everywhere. Due to the adjustable depth of field can be a
  • Inspection system can be variably used for different inspection procedures with changing measuring purposes
  • one or more cameras of an inspection system may be used for underbody inspection of a vehicle passing over them
  • optical deflection elements for deflecting the optical path upwards provided on the underbody of the moving object.
  • an optical deflection element is a mirror.
  • the optical deflection element can be provided with a movement device.
  • the optical field of view of the system of camera and optical deflection element can be extended and on the other hand, the optical deflection element can be placed in a dirt-susceptible position, if no inspections take place.
  • Deflection element with movement device be designed as a horizontally tiltable mirror, outside of the inspection operation, a specular surface of the mirror is tilted down and is pressed only within the inspection operation so upward that an optical path between the camera and a section to be inspected of the moving object arises.
  • a sensor is a continuously measuring sensor.
  • the continuously measuring sensor can continuously collect raw data, for example, by continuously recording an inspection section.
  • the discretely measuring sensor can detect raw data discreetly.
  • the discrete-measuring sensor may collect raw data about the inspection section at one or more predetermined times, such as taking individual images.
  • acoustic / vibratory sensors e.g. a microphone or a vibration sensor.
  • a sonogram for inspection of the inspection section can be recorded.
  • the sonogram can preferably be recorded frequency-specifically and evaluated. This allows, for example,
  • Vibration sensors may be in the driving distance, e.g. in the course of a
  • railway network be provided.
  • an imbalance a weight distribution and / or damage to the vehicle can be measured.
  • Rails for trains are usually well coupled to the train and mechanically strong. If the acoustic or vibratory sensors are arranged on the rail, an exact inspection is possible. Preferably, groups of such acoustic or vibratory sorters are arranged along the rails. Each group is preferably in one integrated sensor module summarized. It is known to record wheel bearing sonagrams. According to embodiments of the invention, it is provided to inspect moving components on the moving vehicle, such as one or more selected from the group consisting of doors, step (esp., Grit in winter), flaps, valves, drive motor bearings, air conditioning and traction motor fan, and create a sonagram for each of these components.
  • Directional microphones can be provided as acoustic sensors. From a sonagram, in embodiments of the inspection method using an algorithm, known noise profiles can be subtracted.
  • One aspect of the invention relates to the acoustic evaluation of a moving train.
  • the inspection system may include a vibration sensor rigidly attached to a side portion of a rail. It is conceivable, in particular, a plurality of
  • sensors in the sensor arrangement are in embodiments one or more olfactory sensors or chemical detectors. They may preferably be used to inspect air conditioning systems of the vehicle that contain refrigerants to inspect the tightness of the vehicle's air conditioning system. Also preferred as olfactory or chemical sensors in the sensor array are smoke sensors and sensors for determining oil leakage.
  • sensors in the sensor arrangement are in embodiments one or more electromagnetic sensors. This is beneficial because people enter
  • the electromagnetic sensor can preferably generate an actual image and the inspection method provides for a comparison of the actual image with a target image or a comparison to abnormalities.
  • the electromagnetic sensor may preferably be used to inspect and detect a short circuit on the vehicle or an electrostatic charge, for example on an air conditioning door. For human personnel such technical conditions can be life-threatening. Therefore, automatic maintenance by a present inspection system according to the invention is particularly advantageous for electrical related technical conditions.
  • a robotic arm can be provided in the inspection system, in particular in order to position the electromagnetic sensor for a near-field measurement on the vehicle.
  • Embodiments Location and / or timer preferably set up for GPS, RFID or WLAN Bearing.
  • timestamps or local punches may preferably be generated for raw data collected by the sensor arrangement during the inspection.
  • the measurement technique of the sensor is fixed.
  • a fixed focus, a fixed aperture and / or a fixed focal length are provided in a camera and the adjustment of the camera to the inspection area via translation and / or rotation of the camera.
  • the camera properties can be kept constant, which can increase the reliability of the camera and simplify the inspection.
  • Triggering Applying the sensor arrangement for acquiring raw data, in particular synchronous recording with a plurality of sensors, in some embodiments comprises triggering at least one sensor and / or storing the raw data sufficiently fast. Triggers may be particularly suitable for inspection sections or inspection objects that have an optical contour that is recognizable / identifiable. Such
  • Inspection objects are preferably antennas, isolators, current bands and
  • Inspection objects in the roof garden of a vehicle can have the advantage that the inspection objects can always be present in raw data in a defined and comparable image position.
  • the inspection object can thus be used within the scope of the
  • Inspection procedure can be detected faster, an extraction from the raw data can faster, and an angle or perspective is determined directly in the raw data, which otherwise would be computationally expensive to calculate or may not be calculable. Triggering also reduces the number of raw data required, eg necessary image recordings, which reduces the volume of data that has to be processed and transmitted. Thus, the significance per bit can increase and computing power become cheaper.
  • more than one defined angle / image position is triggered.
  • 3D information about the inspection area can be obtained, such as a 3D model.
  • An all-round view is also available, a redundancy of raw data, a safety adjustment and a correlation adjustment.
  • Recording by means of triggers can also be faster than untriggered recording, because e.g. the amount of data to be processed decreases. This can be important for the inspection of fast moving vehicles.
  • a light barrier is provided as trigger hardware. It is preferably adapted to a size of the inspection object or inspection section, but nevertheless preferably provides a structurally large object distance.
  • a preferred size of the inspection object is preferably between 5-10 cm. This order of magnitude can be easily determined e.g. covered with just a few pixels of a camera.
  • the light barrier is a laser light barrier. This can be very accurate, especially by a low optical divergence. This is the most accurate way to trigger. Since a light barrier usually requires a defined reflector, which can be provided on the inspection system, in particular an inspection gate, and not on the vehicle, the light barrier is only for lateral triggering of a
  • a camera trigger is provided in some embodiments, preferably for searching the beginning of the train.
  • the associated camera is with yours
  • Image detection direction preferably directed vertically downwards.
  • the camera is preferably provided in the system for inspection of the vehicle anyway. So she can take on another function.
  • a documentation can be created for the purpose of verification that actually a vehicle has passed the inspection system and preferably also the vehicle is identified.
  • measurement data may have been made plausible, for example incorrect, ambiguous or incorrectly assigned.
  • train type or train component type detection or at least categorization may be performed.
  • a type recognition may possibly be carried out from correlation with other data.
  • using the trigger camera preferably a
  • the measurement program may preferably be selected for an ICE2, an ICE3, an ICE4, a train of a particular class, a truck, a cargo ship, an air conditioning type, a pantograph or other component or component of the vehicle.
  • the trigger camera is adapted to a surface image of the
  • color detection takes place on the captured image capture.
  • an already existing color feature such as web red with bleaching stages, or an attached optical marker can be detected.
  • Optical markers on the vehicle may preferably be ICE2 stickers in green, ICE3 stickers in red, a barcode, a QR code, a molecular reflector, which preferably also works when it is partially soiled, or a train number.
  • wavelength-selective illumination and detection can be used in preferred methods.
  • a green light and a green band filter are provided in hardware. This can have the advantage that a sensor with black-and-white capability is sufficient for triggering and a possibly more expensive color camera can be dispensed with.
  • the trigger camera preferably starts a measuring cycle on the moving vehicle by means of a trigger command, that is to say an inspection process with the vehicle
  • an RFID sensor is provided in some embodiments.
  • this can identify a VHF RFID signal, in alternatives a UHF signal.
  • RFID chips are often already on vehicles, e.g. Trains, so that they have a new, additional function in the implementation of the proposed
  • the RFID chip can
  • the inspection system can preferably store an inspection protocol based on the measured data on the chip.
  • a car sort order can be stored in a passenger car.
  • a load control can take place,
  • a light scanner is provided in some embodiments.
  • a light scanner advantageously requires no defined reflector.
  • the vehicle to be inspected serves as a reflector. Since the light scanner can detect changes in the light incidence, it can be particularly suitable for the detection of unknown objects.
  • a light sensor can fulfill the measuring purpose of localizing certain inspection objects, eg in place and time or also location and speed. Advantage may be that of
  • the light scanner is adapted to trigger only selectively in a certain distance range to the inspection object.
  • the distance range can be adjustable, parameterizable or dynamically parameterizable in real time.
  • a dynamic real-time programmable light sensor may first include a current collector, e.g. be detected at a distance of between 1 m to 1, 5 m and then reparameterized to detect an air conditioner at a distance of between 1, 25 m to 2 m.
  • the measuring accuracy of the light sensor may depend on the vehicle speed.
  • the light sensor can preferably point with its optical effective direction from top to bottom, so that contamination can be prevented. It may also have active or passive pollution protection in embodiments, as well as other optical sensors.
  • a non-visible, preferably optical trigger detector is provided in some embodiments. This is preferably used in the passenger area and the inspection is not visible to the passenger.
  • the invisible detector may be an ultrasonic sensor or may receive an optical wavelength not visible to humans, e.g. Infrared.
  • a wavelength-selective filter can be arranged in front of an objective of the invisible trigger detector.
  • the inspection system preferably has a laser source with a first camera at an acute angle to the left and right of the moving vehicle and an obtuse angle to a second camera.
  • the cameras are preferably arranged sequentially in the direction of travel with the smallest possible distance, that is to say with a narrow distance from one another, in order to improve image correlation.
  • the left and right subsystems are triggered simultaneously by the trigger hardware.
  • a preferred frame rate of a camera in the inspection system suitable for use with laser metrology has a frame rate of a maximum of 35,000 frames per second. For improved picture quality, a frame rate of about 5000 fps is used.
  • Cameras are preferably arranged at a distance of 50cm - 100cm and the inspection performed by the cameras at a relative speed of vehicle and inspection system of about 5 km / h. Such a speed is advantageously available on all ICE series over the function "AFB", automatic driving brakes, preset.
  • components are inspected in the roof garden of the train using a camera and laser measuring technology, particularly preferably wind deflectors on the current collector of the train.
  • the inspection process should preferably be set up continuously.
  • continuously measuring sensors or continuously measuring discrete sensors are usually advantageous for air conditioning systems, roof surfaces and the detection of burn-in holes.
  • a frame rate and / or exposure time and / or exposure of the continuously measuring discrete sensor is preferably adaptable, preferably to the vehicle speed.
  • a preferred method comprises a step of providing two or more sensors of the sensor array along and / or orthogonal to a direction of movement of the vehicle, wherein at least one of the sensors is an optical sensor and preferably two or more sensors of the sensor array along and / or orthogonal to the sensor array
  • Movement direction of the vehicle are spaced apart.
  • a first sensor which is upstream of a second sensor in the direction of movement of the vehicle, trigger the second sensor, as described above.
  • the first sensor may alternatively or additionally be spaced orthogonal to the second sensor, that is to say mounted higher or lower than the second sensor. But is preferred in
  • Embodiments that two or more sensors of the sensor array are arranged at the same height.
  • the inspection procedure preferably comprises a step of creating a
  • Diagnosis related to the technical health of the vehicle using raw data acquired by the sensor array The steps of storing raw data, extracted data and / or metadata generated from raw data by the inspection system into a database, which is preferably arranged remotely with respect to the inspection system, and / or applying one are particularly preferred Algorithm on the raw data stored in the database, extracted data and / or metadata, and / or creating the diagnosis that relates to the vehicle, based on the application of the algorithm.
  • all data that exists about a moving vehicle can be stored in a preferably digital and cloud-based data network as a digital record as a "medical record.”
  • metadata are preferably generated from the inspection data, eg, raw data, and in this so-called
  • metadata may be data derived from inspection data, and metadata may be restricted to only one of the
  • the metadata may include weights (eg, on a scale of 0-1) that encode the relevance of a parameter for the inspection purposes.
  • Metadata may also include "diagnoses" or predictions, such as a probability that a pantograph will fail within a given time period, and metadata may include action instructions, such as repairing or exchanging a particular pantograph at a particular time interval.
  • the data network can extract and / or statistically evaluate higher-level metadata algorithmically from medical records of various moving vehicles.
  • These higher level metadata may include, for example, in which
  • Maintenance intervals is a typical pantograph whose train may have to be reversed, serviced or replaced on certain routes.
  • Parent metadata may also comprise a set of weighting factors of parameters containing the information as to which parameters in which weights for the
  • the medical record comprises a train identification.
  • the train identification can, for example, via a data transmission device, such as a
  • Transponder or an example SOFI antenna are transmitted to an ICE train from the train to a data network and / or an inspection system. It is also conceivable, the Glasidenttechnische by an optical image recognition method from a feature of Train, such as the train number on the side or for a train model characteristic design features of the outer shell or previous damage read out. To support and classify train identification, a number of
  • Auxiliary measurements are taken and evaluated by measuring equipment.
  • a measuring apparatus may be a number of photoelectric transducers for measuring characteristics of the exterior shape of a train.
  • a simple photoelectric barrier can determine whether a train or a car of a train, such as a dining car, exceeds a certain height. The height of a turn can significantly limit the possible draw type.
  • a measurement of the external shape of a train is also independent of the pollution or the technical function of the train, other than markings such as a train number on the train, due to
  • a light barrier arranged horizontally can in particular raise the number of wagons and the length of the wagons of a train, which makes it possible with simple means to draw conclusions about the type of train, its series of wagons and / or its orientation with respect to the direction of travel. It is also conceivable that the train actively communicates with an inspection system.
  • the train can actively communicate its identity and other information concerning, for example, its previously determined inspection results. This can be done wirelessly, for example.
  • Exemplary wireless communication technologies are WLAN, radio, RFID, light signals and / or acoustic signals such as a
  • a preferred inspection method comprises comparing present actual data with historical war data and / or expected target data preferably within a data type, the data types comprising raw data, extracted data and / or metadata with historical raw data, extracted data and / or metadata, and creating diagnosing the vehicle by determining deviations of the compared data.
  • Data processing enables, for example, an automated and supraregional
  • inspection data can be collected from an inspection system in the Munich workshop by an inspection system become.
  • This inspection data can be processed into a diagnosis by comparing it with a centralized database of empirical and metadata data, for example, that a coffee machine or air conditioner will be defective by 60 percent within the next week. Preserving the exact type of forecasted defective air conditioning in each factory workshop would require expensive inventory.
  • Replacement air conditioning system of the right type is stored in Cologne and could be delivered to Dortmund by a forwarding agent in the operating workshop Dortmund, which the train timetable happens, in less than the travel time of the train to Dortmund),
  • an inspection method with at least one of the steps of recording raw data by the sensor arrangement and / or buffering of raw data on a first storage medium, preferably within the
  • Sensor arrangement is arranged, and / or after completion of the recording of, preferably all, raw data, analysis of the raw data and deriving extracted data based on the analysis and / or collection of metadata from raw data and / or extracted data and / or archiving of raw data and / or extracted data and / or metadata on a second storage medium, wherein the first storage medium has a higher writing and / or reading speed than comprising the second storage medium, and preferably remote from the sensor assembly, and / or correlating raw data, extracted data, and / or metadata using the algorithm for generating the diagnosis and / or storing the diagnosis in a database, preferably associated with a vehicle the raw data was collected.
  • an inspection method including at least one of deriving an inspection result for the vehicle based on the diagnosis, preferably confirming the inspection result by a user, and preferably giving a maintenance recommendation to the user and / or proposing a user
  • the inspection method includes a step of generating big data using the inspection result.
  • big data or “mass data” primarily refers to the purposeful and purposeful processing and evaluation of large, complex and rapidly changing ones
  • Amounts of data The global data volume is estimated to double every two years. This development is mainly driven by the increasing machine generation of data z. Eg via protocols of telecommunication connections (call detail record, CDR) and web access (log files), automatic capturing of, for example, RFID readers, cameras, microphones and other sensors.
  • CDR detail record
  • log files web access
  • the concept of "Big Data” is finding its way into more and more areas of life, such as the automation of production processes (Industry 4.0, Internet of Things) .In accordance with this concept, the concept also becomes intransparent Automation of decision-making processes through the use of software and hardware.
  • the concept of "big data” requires not only implementation in software but also hardware such as the appropriate communicative violation of data processing equipment or at least partially automated inspection systems, maintenance facilities, computer systems for planning the use of scarce resources (such as the occupancy of maintenance facilities or tracks for Trains) and / or Warehouse or Logistic Systems
  • the present invention is particularly concerned with approaches such as
  • Data mining refers to the systematic application of statistical methods to large volumes of data (in particular "big data”).
  • Mass data with the goal of detecting new correlations, dependencies or trends. Due to the amount of data, the complexity and the computational effort, such methods are manually or partially manually economically and or in real time not feasible.
  • Data Mining is sometimes defined as a subordinate term to the entire process of knowledge discovery in databases (KDD), which also includes steps such as data collection or preprocessing, while data mining is actually just the process of parsing the process
  • Pre-processing for example, data cleansing that integrates sources and removes inconsistencies, for example, by removing or
  • Transform for example, transformation into the appropriate format for the analysis step, for example by selecting attributes or discretizing the values;
  • the tasks of data mining can be:
  • Cluster analysis grouping or correlation of objects, properties of
  • association analysis identification of relationships and dependencies in the data in the form of rules such as "A and B usually follow C";
  • Methods of data mining are used according to the invention, with the least possible use of resources on hardware (for data processing, for inspection, maintenance, storage and / or other infrastructure) and software
  • outlier detection looks for data objects that are inconsistent with the rest of the data, such as having unusual attribute values or deviating from a general trend.
  • a "density-based algorithm” For example, outlier detection looks for data objects that are inconsistent with the rest of the data, such as having unusual attribute values or deviating from a general trend.
  • Outlier detection for example, identifies objects that have a density distinctly different from their neighbors, where the density can be related to the number of errors per unit time of an object, and then identifies outliers manually and after a negative plausibility test be hidden from the record in order not to distort or degrade the heuristic of the metadata in the sense of the pursued purpose.
  • Cluster analysis identifies groups of objects that are somewhat more similar than other groups. Often these are clusters or "clusters" in the data space, examples of densely clustered clustering techniques in which the clusters can take any form are DBSCAN or OPTICS, but methods such as an EM algorithm or k-means algorithm are most commonly used in the
  • Dataroom spherical clusters for application. Objects that have not been assigned to a cluster can be interpreted as outliers in the sense of the aforementioned outlier detection.
  • the cluster analysis identifies groups of objects that often only have to sample them based on a sample, which is the number of
  • the Demand planning of inspection systems and inspection and / or maintenance intervals can be planned more efficiently.
  • Classification like cluster analysis, is about classifying objects into classes.
  • the classes are generally predefined (for example: bicycles, cars, trains or types of trains).
  • Previously assigned to any class objects are automatically assigned a class.
  • existing classes are not expanded, for example, by changing the criteria for class membership. This would not be relevant to practice, for example, existing train types and would also with an increased computing and
  • Metad Atensatz A may contain the information that wheel bearings of a train have an above-average defect rate
  • Metadata set B can tell me which trains have passed a certain stretch of road in a given period of time. The final rule would say that trains to which metadata sets A and B refer, have a higher probability of defective wheel bearings than other trains traveling on other sections.
  • Regression analysis allows redundant information to be removed, reducing the complexity of the data.
  • a reduction in complexity enables targeted intervention in a highly complex system such as a rail network with targeted means that have an understandable, predictable or predictable effect.
  • learning algorithms reveal correlations in datasets, without necessarily providing an analytical understanding of the causal relationships of these correlations.
  • temporal correlations play an important role.
  • time series analysis it is possible, for example by means of special distance functions such as the dynamic time warping distance, to recognize similar characteristics in temporal progressions of different data sets, even if the data records were determined at different times.
  • Studies can be provided, the results of which can be compared with the automated results.
  • the parameters of the automated method can be adapted in order to bring the results of the automated method at least with partial results of the regular investigations into coincidence.
  • a virtual mirror system of the actual system (for example of a railroad network) can further be created in which changes in the parameters and their favorable or unfavorable effects on the overall system can be simulated, tested and predicted.
  • it may be provided to classify changes to the overall system and to carry out a simulation of the consequences on the overall system for each change from certain classes.
  • This is particularly advantageous for highly complex and highly safety-relevant systems such as railway networks.
  • a procedure as shown in the previous section has in particular the
  • the method becomes more efficient by means of statistical data evaluation, in that learning algorithms preferably document empirical values in as compact a metadata as possible with regard to the data volume.
  • this metadata can be called
  • Weighting factors for inspection parameters for example: missing screw is a must for immediate maintenance
  • selection factors for inspection purpose relevant inspection measurements for example: for the three-dimensional inspection of a pantograph of a train one or two optical cameras are sufficient
  • the invention provides for a human controller, as it were a "pilot", who, on the basis of his experience, the knowledge of superior goals and due to
  • Storage system are recorded but can be ordered easily and quickly on the open market) and / or prioritization (for example, it is less important, a not
  • the cost sequence of such interventions is preferably estimated or predicted, so that the involved parties are ideally informed about the cost consequences of such interventions on the basis of a concrete cost estimate.
  • This technical implementation creates a marketplace for the parties involved, in which they can transparently, quickly (ideally in real time) and, while ensuring the security of the overall system, align their goals and balance the costs. Without a technical implementation and networking of a complex system such as a railway network, this would not be possible.
  • the operator of a train fleet receives immediate feedback, which additional costs, for example, of the workshops is to be expected if a particular train is to be treated during maintenance preferred to comply with the timetable.
  • a workshop also receives the feedback as to which follow-up costs would result from a non-early repair of a train due to the delay or the failure of connecting trains. This is how a workshop manager can do
  • the intervention can also cost other parties and intervene in their interests and competences.
  • the intervention of a specific party in certain parameters or goals is subject to competence define. These areas of competence may relate, for example, to a changeable parameter range or to the cost ensuing therefrom for the party or another party.
  • a veto system can be implemented, in which another party can contradict the intervention of a first party when a defined cost center is exceeded.
  • the algorithm can also determine a benefit for another party through the intervention of a first party.
  • the algorithm may propose a cost distribution and, in particular, after an optional release step, initiate resource scheduling on this basis.
  • the costs that underlie the algorithmic optimization of the rail network and the actual costs, which the involved parties have to bear economically, can also fall apart retroactively.
  • the resources of, for example, a rail network can be arranged quickly and smoothly in terms of the overall system, because the balance of particular interests can be downstream.
  • an operator of a train fleet will be charged for the repair costs for the urgent maintenance of a particular train and the opportunity costs for its breakdown.
  • An operator of a maintenance facility is shown that he can better utilize his maintenance resources through the unscheduled and urgent repair.
  • An exemplary optimization result is that a probably necessary repair on a train could be carried out at a first operating workshop with a more favorable total cost due to the current utilization than in a second company workshop, taking into account the logistics costs.
  • the method according to the invention objectifies, speeds up and, moreover, provides greater planning security between the different parties involved.
  • each party has a "pilot" with manual intervention rights in the automated optimization.
  • a pilot can be an individual or a group of people. This way, a multi-user principle for security-relevant or cost-relevant decisions can be made in an efficient and secure manner
  • Exemplary adaptive algorithms which are suitable for the method according to the invention are Monte Carlo algorithms, traveling salesman algorithms, neural networks or evolutionary algorithms.
  • learning algorithms can also be used according to the invention. Exemplary implementations and concepts of such adaptive image recognition algorithms are described in the seminar series "Technologies of Factory Automation: Learning Machine Vision for Quality Assurance Based on Supervised Learning: Support Vector Machine” by Patrick Raulfug,
  • a preferred inspection method comprises at least one of the steps reading in a maintenance plan of all vehicles and / or recognizing whether a vehicle is the
  • Sensor arrangement and / or performing a functional test of the inspection system to detect in particular non-functional sensors, and, preferably based on automatic detection of a functionality of the inspection system and / or selecting an inspection method based on historical data depending on the vehicle to be inspected and / or determining
  • Environmental parameters preferably an ambient temperature, before or after
  • the inspection method retrieve a maintenance plan that contains information about which train operates at what time. Further information may refer to the composition of the train, such as number of cars or
  • Wagon series or type of wagons e.g. on trains.
  • Wagon series or type of wagons e.g. on trains.
  • Inspection Procedure provides RFID detection on an approaching train to alert the inspection system that will be traversed next by the train.
  • the Pre-warning then preferably initializes the inspection system, in particular
  • Sensor arrangement preferably camera, and lighting system. Thereafter, a functional test is preferably carried out, if all components of the inspection system have been initialized. Thereupon, a feedback is generated as to whether or not the inspection system is functioning. Thereafter, preferably, the detection by the
  • Inspection area of the vehicle to inspect.
  • a specific inspection program is provided for each type of train, type of wagon and / or wagon assembly.
  • some train types have elements that are missing from another train type.
  • An inspection can skip the inspection of these items if they are not there anyway.
  • the acquired raw data is preferably first displayed on the sensor, e.g. the camera, cached locally.
  • the raw data is then, preferably with a time stamp, unprocessed to a
  • Edit memory also called analysis hard disk sent. That's how it works
  • Data volume can be reduced.
  • black and white images can be generated by the sensor array, which can bring a reduction by a factor of 1, 5 to 3 compared to color photographs.
  • black-and-white images can be sharper than the color image and provide better contrast, especially since no color interpolation, for example by a Bayer matrix, is necessary.
  • the lighting can be reduced.
  • 3.1.16 Noninvasive Measurement Technique applying the sensor arrangement with non-invasive measurement technology from the outside, based on the vehicle, wherein preferably a distance between sensor and vehicle, preferably during the acquisition of raw data, at least temporarily between 0 cm and 250 cm, preferably between 10 cm and 100 cm.
  • the distance to the object to be inspected is, in particular in the case of optical sensors, after a weighing optimization of the number of pixels to be at least resolved and of the image section to which the
  • Inspection object should fit as completely as possible, preferably during a
  • Inspection object of e.g. 1 mm to 1 cm, at least 3, preferably 5, pixels (or pixels) are to be resolved.
  • the inspection method comprises applying the invasive sensing sensor array externally to the vehicle and / or applying the sensor array, wherein at least one sensor of the sensor array is mounted to the vehicle for measurements in or on the vehicle whose raw data and / or metadata generated therefrom are transmitted away from the vehicle.
  • these are one or more markers from the group consisting of barcode, strip, train number, beacon, RFID chip and WLAN bearing.
  • a preferred inspection method comprises the step of generating a three-dimensional model of the inspection section from one- or two-dimensional images of the inspection section detected by one or more sensors. It is preferably provided, three-dimensionally resolved data sets of
  • the inspection method preferably derives a three-dimensional model of the inspection section from a singular or a plurality of three-dimensionally resolved data sets from one or more sensors or sensor types. Namely, according to another idea of the invention, a three-dimensional model of at least one inspection section of the vehicle to be inspected can be generated even with only one imaging sensor such as a camera.
  • the approach according to the invention is to utilize the relative movement of the vehicle to be inspected against the imaging sensor.
  • the imaging sensor is translationally and / or rotationally stationary, which facilitates its location, actuation, adjustment and calibration, and the vehicle to be inspected is moved.
  • One or two-dimensional measurement data of the imaging sensor recorded at different times are thus recorded from different angles with respect to the optical axis of the imaging sensor. From this, a three-dimensional image of at least one inspection section of the vehicle to be inspected can be created by an image-evaluating algorithm.
  • a three-dimensional plan view of a pantograph on the roof of a moving train can be created from which it is possible to check with higher interpretation reliability and also by an algorithm as to whether the pantograph is damaged in its structure, for example broken.
  • groups of technical components on a moving train in particular on the roof garden of a moving train, can be grouped together in clusters.
  • clustering such components pattern recognition from image data can be more efficient and less computationally and / or
  • a target-actual comparison can be performed more efficiently, in which not the individual components are compared, but the entire cluster with a database of possible target clusters.
  • a cluster for example, a pantograph, an air conditioner, a roof flap and / or an antenna of a train can be summarized. This increase in efficiency results in a lower data volume with the same inspection accuracy. At the same sampling rate of the inspection system can thereby, for example, the relative speed of the moving object and
  • the efficiency of the inspection process depends on the appropriate choice of parameters and, in particular, input parameters. These parameters must be identified and weighted for each technical and economic system. By suitable preselection of the parameters to be taken into account can not only an automatic
  • An exemplary method according to the invention is implemented between a diagnostic system, a storage system and a repair system. Between the systems flows of goods and information flows are exchanged.
  • the method may include the following steps:
  • the inspection procedure can be used for inspection of a variety of technical
  • Functional states of the vehicle are used, preferably for crack detection or for detection of burn-in holes or for inspection of one or more elements from the group consisting of covers, antennas, air conditioning grilles,
  • the inspection method may preferably be used to inspect the technical condition of covers on the vehicle.
  • shape matching is performed on a central strip of a roof on which covers may be located of the vehicle.
  • the parameters used are preferably edges of outer edges of the covers and / or a respective outline of screws of the cover. Through appropriate parameter selection in the analysis of many trains edges could be determined, which has every cover and which are distinctive for this type.
  • a variation model is preferably used. It may have taught-in examples of covers.
  • Shape matching is the process of looking for objects in a scene or inspection section using a contour model, for example in a roof garden of a train.
  • This contour model contains edge information and searches for these edges in the scene using a variety of techniques. If the agreement is sufficiently high, the inspection object applies or the inspection section found.
  • the accuracy and consequently the speed of the shape matching process can be regulated by various parameters.
  • the inspection method may preferably be used to inspect the technical condition of antennas on the vehicle.
  • shape matching may be performed on a selected area in the center of the roof in which the antennas may be located, eg with a camera. It is preferable to select different antenna types from above from this perspective.
  • edges of the outer shape of the various antennas come into question.
  • a lateral lateral shape model for example in a roof garden of a train.
  • This contour model contains edge information and searches for these edges in the scene using a variety of techniques. If the agreement is sufficiently high, the inspection object applies or the inspection section found.
  • the parameters used are preferably one or more elements from the group consisting of the width of the antenna, height of the antenna, blob analysis, results of shape matching, main axis transformation and moments of the image (center of gravity, mean, etc.).
  • the model for the matching methods was created on the basis of many test series with a large number of antennas.
  • an outline of the antennas is preferably measured.
  • a line method and / or a comparison with good parts can be carried out.
  • a light / dark threshold can be used in the identification of the antenna.
  • the presence of nuts and bolts can be determined by classifiers.
  • the parameters used are preferably a threshold value of the threshold method and / or a number of irregular lines and damages. Errors are then clearly indicated by the aforementioned measurements due to differences in brightness and / or differences in shape.
  • the inspection method may preferably be used for inspecting the technical condition of components of the vehicle with regard to cracks. Contiguous lines (brightness rating) Threshholding to detect irregularities
  • the candidates are classified. Suitable parameters are the size and shape of the extracts and / or brightness differences in the area of the hole candidate. A preferred
  • Parameter selection with regard to the size of the cracks depends on the specifications.
  • Thresholds for cracks are preferably chosen on the basis of typical patterns for cracks that have been identified by analysis of many features. For example, the presence of cracks on a component of the vehicle can be inspected as a technical condition.
  • the inspection method may preferably be used for inspecting the technical condition of components of the vehicle with regard to penetration holes.
  • a blob analysis using Threshholding is used to find potential candidates. Subsequently, the candidates can be classified and evaluated. Suitable parameters are the size and shape of the extracts and / or
  • Brightness differences in the area of the hole candidate When selecting the parameters, the size of the holes depends on the specifications. A threshold was chosen based on typical patterns of cracks detected by analyzing many trains.
  • the inspection method may preferably be used to inspect the technical condition of air conditioners of the vehicle.
  • dark areas are searched for in possible areas, particularly preferably from a perspective from above onto the vehicle.
  • the parameters chosen are preferably shape and / or size.
  • lateral grids are inspected.
  • a center of the lateral grid is first searched.
  • blob analysis a form of the grid is then cut out.
  • the threshold for dark areas is selected for a range that has been determined by analysis of many recordings.
  • Novelties irregularities
  • a Fourier transform is performed to check a frequency of the grating.
  • Thresholds for imperfections may increase robustness in embodiments of the inspection method and avoid detecting slight bends.
  • a number of defects or thresholds for detecting errors can be used.
  • Parameter selection is based on the size of the grid pattern.
  • At the inspection of the technical conditions of climate curtains from the side are basically filtered out parts in the grid, which do not belong to the structure. Novelties are searched for, as well as in the inspection from above.
  • Quick-release fasteners can be inspected for their presence and an angle of notch, preferably by blob analysis,
  • the parameters used are preferably the size of the grid pattern, the size of errors and / or a number of quick-release closures.
  • the choice of parameters is preferably based on specifications from the specifications and a validation by tests on many vehicles. Thus, preferably, the presence of the climate grid or its integrity as a technical condition can be determined.
  • the inspection method may preferably be used to inspect the technical condition of air conditioning louvers of the vehicle.
  • a shape matching is preferably used in terms of a lamella shape.
  • a blob analysis of the interstices between the climate lamellae is performed. A number of undamaged ones
  • Environmental lamellae can be determined using a line measurement method.
  • the number of slats is predetermined and their distance is always the same.
  • the parameter selection is thus simplified.
  • the presence of a Klimalamelle or their integrity as a technical condition can be determined.
  • the inspection method may preferably be used to inspect the technical condition of screws, in particular screws of an air conditioning system or in a roof garden of the vehicle.
  • a surface with screws preferably on the air conditioner, is searched. Then done
  • Brightness differences from a schematic black and white image of a screw as it can minimize errors due to contamination. For example, you can
  • Screws are inspected at a pantograph of the vehicle, preferably from below. Then the inspection system can have one or more mirrors and one
  • Screws is unique, a triggered inspection can be done. Generally speaking, distinctive locations in the inspection section can be searched for through thresholding and position-dependent locations with screws can be found. The screws can then be classified as "present” or “absent”, preferably based on the shape and gray scale of the extract and / or brightness differences. Classification parameters, In particular, to differentiate screws from dirt can be done on the basis of a broad analysis of all the screws of several vehicles, such as trains. So can
  • the presence of a screw of the air conditioner can be determined as a technical condition.
  • the inspection method can preferably be used for inspecting the technical condition of lines of the vehicle, preferably for electrical lines and / or fluid lines.
  • Shape Matching using the shape of the leads and the shape of screws and objects near the leads. In particular, the position, the edges of the screws and a classifier for
  • the parameter is a threshold value of the threshold method and a number of irregularities and lines.
  • the choice of parameters is made by tests of typical impurities and a difference to errors. Through many images, a threshold can be evaluated. For electrical lines, ie cables, a shape matching with
  • Air ducts can shape matching based on the shape of the air ducts.
  • a contrast-rich model for the separation of other components preferably the roof garden of a train, preferably used.
  • An optimal model for the inspection can be determined after analyzing several moves. It is particularly preferable to inspect whether a connection from a point A to a point B formed by the line is intact. This can be done by line method or a course through
  • Brightness information is done.
  • interrupts are detected.
  • Length, curvature and / or position of the conduit are also detected in some embodiments.
  • the parameters are preferably the line shape, a number of lines and a length of lines.
  • the connection is best measured by a coherent line.
  • the inspection method may preferably be used to inspect the technical condition of a main switch of the vehicle.
  • the main switch is usually attached to a main line. Based on the contour of the position of the control element of the main switch can be inspected with shape matching, for example, by determining a certain position relative to the body of the main switch. Also, screws of the main switch can be classified and current bands undergo a texture classification. The texture on the current bands is unique and based on tests, the
  • the presence of the main switch, its position or its integrity as a technical condition can be determined.
  • the inspection method may preferably be used to inspect the technical condition of insulators on the vehicle.
  • the inspection is carried out
  • an inspection can be made from the side by shape matching on triggered images, since a slight shift despite triggers is possible. Triggering is preferred because isolator positions are always unique, but due to strong occlusion, only a few positions are possible where each isolator can be seen. Insulators can also be inspected at an angle. For this purpose, it is preferable to search for screws and shape matching using the lamellae. A triggered measurement is preferred for oblique inspection. If the inspection is made from above, a circle above the center is preferably determined on the surface of the insulator. Damage is over
  • Thresholding can be determined. Screws can be inspected by shape matching the contour. If the inspection is carried out from the side, slats are preferably sought. For this a line method can be used. Brightness transitions are preferably determined and / or a parallelism of the slats evaluated. It is preferred that the number of slats is counted. In particular, interrupted lamellae can be identified.
  • the parameters used are preferably a threshold value of the thresholding, a number of lamellae, the size and shape of the lamellae and / or a brightness discrepancy in lamellae. The choice of parameters may preferably be made by testing many vehicles and analyzing individual isolators on each vehicle.
  • the inspection method may preferably be used to inspect the technical condition of current bands at a pantograph of the vehicle.
  • an optical flow method for better separation can be used, because the contrast between the current band and other components in the inspection section is not very large.
  • the background can be separated from the object by means of optical flow, thus creating a rough height map of a scene. This allows the background to be separated from relevant objects.
  • the presence of a compound from A to B is inspected. This can be done by detecting a disparity on two different images or by determining height information.
  • the surface of the Current bands are analyzed in embodiments by a texture model where
  • Differences in the texture model can be determined.
  • cracks are determined from 20% of the material width.
  • the parameters used are the width of the current band, the length of the current band and a number of abnormalities on the band.
  • the texture on the Stromband is almost unique and errors in the texture are easily recognizable.
  • the inspection method may preferably be used to inspect the technical condition of a circuit breaker of the vehicle.
  • a lever of the vehicle Preferably, a lever of the vehicle
  • Disconnectors can preferably be detected and inspected with NCC (normalized cross correction) light / dark.
  • NCC normalized cross correction
  • an angle of the lever to the horizon is determined.
  • the screws can also be inspected via shape matching and correlation-based matching over the contour.
  • Parameters are the number of screws and the position of the lever suitable.
  • the inspection method may preferably be used to inspect the technical condition of a converter on the vehicle.
  • the surface of the transducer can be inspected for cracks and / or burn-in holes, preferably as described above.
  • the inspection method may preferably be used to inspect the technical condition of a bellows on the vehicle.
  • a shape matching of the shape of the bellows takes place. Since a bellows has a unique shape, edge-based
  • Cracks in the bellows can be determined by line method. Parameters can be the number and size of the damages.
  • the inspection method may preferably be used to inspect the technical condition of a pantograph linkage on the vehicle.
  • a pantograph linkage on the vehicle Preferably, a
  • Pantograph linkage searched for irregularities.
  • the parameters include the number and size of the fault locations, the threshold value of the thresholding method and the number of Rods suitable.
  • the number of rods is clearly measurable.
  • the detected bugs indicate the severity of the damage.
  • the threshold can be evaluated by tests.
  • the technical condition of a pantograph joint is inspected. For this purpose, preferably lines are determined and angles of the extracted lines are calculated. Damage is transferred in embodiments of the inspection process
  • Brightness analysis and difference method are detected. Suitable parameters are preferably the tolerance at angles and the size of damages and scratches. About a wrong angle can be concluded on error. The size of the detected error indicates the severity of the damage.
  • the inspection method may preferably be used to inspect the technical condition of a pantograph on the vehicle. Preferably, this is done by a height measurement by laser profiles. A 3D image generated by a laser beam
  • Pantograph rocker can be normalized for this purpose.
  • a skew between pantograph headers is determined.
  • the presence can be determined via thresholding.
  • a width of wind deflectors is determined in embodiments.
  • an angle relationship becomes one
  • the geometry of the pantograph rocker is determined. Blob analysis and Linen method can be used to find individual inspection objects of the pantograph rocker, as well as the analysis of brightness discrepancies for measuring individual ones
  • the inspection method may preferably be used to inspect the technical condition of a lifting drive cable of the vehicle.
  • a line is preferably approximated.
  • a metrology model is used in embodiments.
  • the metrology model preferably describes a rough shape in which a
  • the parameters are preferably the shape and length of the rope.
  • the inspection method may preferably be used to inspect the technical condition of a support bracket on the vehicle. Shape matching can be used to detect the shape of the support bracket. With the help of a found model an angle can be calculated. The angle is set as parameter by default.
  • the inspection method may preferably be used to inspect the technical condition of a test cock on the vehicle.
  • a position of the test cock and preferably an associated cable can be determined with classifier.
  • Brightness differences suitable means As a parameter, the presence of the test cock, the position of the test cock or the presence of cable ties can be selected.
  • the parameters are selected according to specifications in the specifications.
  • the inspection method may preferably be used to inspect the technical condition of car transitions of the vehicle.
  • a vehicle preferably a motorcycle, a bicycle, a motorcycle, a bicycle,
  • Gray value projection for light / dark detection The greyscale projection projects a 2D image onto a 1D line. For this purpose, the gray values of the individual lines are averaged and entered in each case in the corresponding pixels of the line. This can be partially valuable
  • Wagon crossing to be inspected This is preferably done again via shape matching.
  • Car transitions can be inspected from above and / or laterally. From above, inspection sections are preferably inspected with thresholding, lines are created, directions of the lines are checked and windings are counted. From the side, in the case of
  • Surfaces can be inspected at car transitions, for example by threshing.
  • bright regions are recognized and the size of the regions determined.
  • a threshold threshold value, a number of turns, and the shape and position of cables are considered parameters. The parameter selection occurs after tests for evaluating thresholds.
  • Data processing system and the inspection system are adapted to perform the inspection method according to the invention. It is preferred that the data processing system has at least one subsystem which is arranged internally relative to the inspection system. For example, this may be a server in an internal data network associated with the inspection system. It is preferred that the data processing system has at least one subsystem which is arranged externally relative to the inspection system. For example, this may be a cloud computing farm in an external data network that is outside the inspection system.
  • the computing power of an external subsystem can be used without completely transferring the measurement data and derived data of the inspection object collected by the inspection system to the external subsystem. This reduces the risk of losing control over the measurement data.
  • the data processing system has software resources and / or hardware resources that are responsive to a transferable and storable data rate for applying the sensor array to the computer
  • Relative speed is preferably between 2 km / h up to 400 km / h, more preferably between 2 km / h and 50 km / h, more preferably between 3 km / h and 10 km / h and ideally at 5 km / h.
  • the data rate is adapted to a constant or variable relative speed during an inspection interval, preferably dynamically to achieve the sufficient and / or constant data redundancy.
  • sufficient data redundancy is at least one, preferably 3 and more preferably 3 to 10 data sets per section of an inspection object to be inspected.
  • Devices and methods according to the invention are, as stated, suitable for vehicles which move at a relative speed to the sensor arrangement. For example, a vehicle can be guided past a stationary inspection system.
  • both a vehicle to be inspected and the inspection systems according to the invention are moved at least intermittently and / or in sections for their detection with respect to a stationary reference system such as an operating workshop.
  • a stationary reference system such as an operating workshop.
  • a train at 5 km / h could travel through an operating workshop and undergo an inspection system with a rectified speed of 2 km / h over a certain length.
  • the relative speed over this length between train and inspection system would be only 3 km / h, so that a higher density of image data and thus a more precise detection over a portion of the vehicle can be achieved.
  • the inspection system it is preferred to arrange the inspection system stationary in order to avoid a more complex and harder to set mobile inspection system. Furthermore, a movable inspection system would have to be moved back to a starting position in a workshop after a maximum travel, so that no continuous inspection operation is possible.
  • the background is that given imaging hardware and a specified frame rate and / or resolution of the images to be captured limits the maximum relative velocity of the moving body.
  • An inspection method is preferred in which the speed, for example, of a train to be inspected is modulated in time. In time phases in which an inspection section of the train is in the field of vision of the sensors, preferably optical sensors, the train can be slowed down and / or travel slower in order to achieve a higher detection rate.
  • the train can be accelerated to speed up the inspection process and increase the throughput and utilization of an inspection system. It is conceivable that for more precise synchronization between the movement of the train and the control of the detectors of the train and the inspection system via a
  • the train may be remotely controlled during the application of the inspection procedure.
  • a remote control device can be provided in the inspection system.
  • Hardware resources provide a data transfer rate of up to 7 Gb / s between a server and the sensor array.
  • the data transmission can be split over several data cables. For example, over 7 copper coax cables per 1 Gbit / s can be transferred to the
  • Camera detector the part of a camera as a sensor, image capturing initially on the camera chip. The images are then from the camera chip over a
  • Data line with preferably at least 300 Mbit / s forwarded. It is preferred that 3 sensors, e.g. Cameras connected to a common switch. The switch is thus set up to receive a data volume from the sensors
  • a preferred switch provides power to the sensors connected to it, either via the data line to the sensors or via a separate power line.
  • An advantage of bundling multiple sensors on a switch may be that the plurality of cables leaving the multiple sensors need only pass the switch, and there may be a reduced number of cables behind it, e.g. only one single data cable connecting the switch to other components of the
  • SSD Solid State Disk
  • Such SSDs may also have a data rate of z. 7 times 1 Gbit / s process.
  • the SSDs can also be redundant, preferably also in a RAID array, e.g. RAID 0 or 1, but it is preferred that the SSDs be operated in a single mode.
  • a separate processor or processor cluster is provided for the data processing of each inspection section. This way you can avoid that
  • a processor or processor cluster exclusively evaluates the roof garden or the pantograph.
  • Inspection of a carriage transition is preferred to extract wagons for wagons from the raw data.
  • image by image of each inspection object is extracted.
  • Preferred methods are to predetermine, for example, none on soft cars
  • the relevant image section is extracted, whereby fewer data and objects are to be evaluated. So can an amount of data to be evaluated for example, from 40GB to 5GB. By using triggers, certain sensors, preferably cameras, can already reduce the amount of data from 40GB to 26GB.
  • the image sections recognized as relevant are stored, preferably on an SSD, as described above. In a further step, in some embodiments, the extracted data becomes one
  • Storage hard disk that is slower than the SSD, such as a mechanical hard drive.
  • the archival hard disk is preferably redundant in the RAI D composite, e.g. as RAID1, RAID3, RAID5 or RAIDI O.
  • Storage hard disk has a capacity of 3 - 6 TB, preferably 5 TB.
  • an interface be provided by camera clients to servers.
  • a user at a camera client of the inspection system can access a server that manages the inspection system.
  • a preferred server is remotely located relative to the sensor assembly. This means that, for example, the sensor arrangement can be arranged in a tunnel while the server is accommodated in a remote data center.
  • An interface for a web interface is preferably provided on the server.
  • the server can be accessed via the Internet via a browser.
  • Another preferred interface is provided between the server and control of the inspection system.
  • the inspection system controls lighting systems and triggers of the inspection system.
  • the lighting and the triggering of the inspection system can take place remotely through the server.
  • the server has an interface to external databases.
  • Such is preferably SOFIS to retrieve train arrival and train type, e.g. to read in a maintenance plan.
  • the server can send control signals to the sensors, e.g. initialize the cameras.
  • the cameras are initialized about 30 seconds before the train arrives.
  • the server may preferably initialize a train search camera and, if necessary, activate the lighting of the inspection system for the inspection sections.
  • Inspection system are preferably provided two types of clients: recording clients and user clients.
  • Recording clients are preferably extra computers for receiving and linking external data without disturbing a calculation computer. This achieves a hardware-based separation between external data reception and processing of measured data.
  • Userclients allow influencing the
  • all hardware resources of the inspection system are connected to a common electrical ground.
  • the common electrical ground is preferably connected to a railway ground. It is particularly preferred that copper strands are provided with a cross section of 50-90 mm 2 between the components of the inspection system in order to derive high voltage can. A particularly preferred
  • Cross section is 70 mm 2 .
  • a surge arrester is provided which causes the inspection system to shut down as soon as the system fails
  • Overvoltage arrester is triggered.
  • a high-voltage protection may be provided, in particular for optical sensors, such as cameras, or acoustic sensors, such as microphones, in particular directional microphones.
  • microphones are lined on housing inner sides with sound insulation.
  • lines between components of the inspection system preferably shielded, copper cables are preferred. You can accomplish data and / or power transmission.
  • a common data and power line between components of the inspection system is provided.
  • glass or plastic fiber cables can be provided.
  • the advantage is that these glass or plastic fiber cables can not conduct electricity. Therefore, glass / plastic fiber cables can be used for the electrical decoupling of components of the inspection system.
  • a glass or plastic fiber cable is connected via opto-couplers at their ends to components of the inspection system for data transmission.
  • the advantage is that plastic is not amorphous, hard to break and easy to assemble.
  • the EM susceptibility is lower than copper cables.
  • the inspection system may include a multi-part inspection port.
  • the inspection system has a sensor module with one or more sensors of the sensor arrangement, which is galvanically decoupled.
  • the sensor system has a first ground connection and a second one Earthing connection on. The sensor module is connected to the first ground connection and at least one other component of the inspection system is connected to the second
  • At least one other component is a trigger line, which is preferably an optical trigger line, such as a trigger line. a fiber optic cable.
  • Trigger line is preferably coupled via an optocoupler with the sensor module to trigger one or more sensors of the sensor module, preferably at least one camera.
  • Automation of an inspection and / or maintenance procedure of a moving object such as a train or rail network has a significant added value compared to manual control, especially if it functions so reliably and smoothly that the inspection system and / or maintenance system itself has little manual controls or Maintenance interventions needed. This is especially true at the
  • any of the apparatus such as a
  • Inspection system an inspection sensor of an inspection system
  • Lighting system of the inspection system a cleaning device of the
  • part of the protocol transmitted to the data network includes metadata. In the simplest case is one
  • Metadata record a preferably periodically repeated signal that the equipment is working properly.
  • one or more apparatuses of an inspection and / or maintenance system can be sensors for measuring
  • Environmental influences include. Environmental influences, for example, an ambient temperature, a humidity of the environment, ambient light,
  • the operating conditions of an apparatus can be a temperature range of -10 ° C to 50 ° C and a maximum
  • measurements of the environmental influences can be an input variable in the evaluation of raw inspection data. For example, if an ambient temperature is below freezing point of water, the increased likelihood may be implemented in the evaluation algorithm that artifacts in the inspection image data are due to icing.
  • Environmental conditions of an outdoor area through which the train has traveled are taken into account.
  • a train is on the free route
  • all or at least selected aggregates of a train can each be equipped with a computer or with a computer which is at least groupwise common to theirs
  • FIG. 1 shows a schematic structure of an inspection system according to FIG.
  • FIG. 1 shows a schematic structure of an inspection system 1 according to FIG.
  • the inspection system 1 provides a passage for a vehicle (not shown).
  • the passage and the exact structural configuration of the inspection system 1 relative to the vehicle is not shown in this schematic view for the sake of simplicity. In this embodiment, this is
  • Inspection system 1 set up to inspect a moving train as it passes the passage.
  • the train is an electrically powered train with
  • Power is supplied by a power line.
  • the high voltage line runs through the passage.
  • the inspection system 1 has a sensor arrangement. Sensors 2a-g are combined in a first sensor module 3 and a second sensor module 4.
  • the inspection system 1 includes a server 5 constituting the data processing system.
  • the server 5 comprises a database for storing raw data, extracted data and metadata acquired by the sensors 2a-g of the inspection system 1
  • Embodiment in a folder structure is arranged to apply an algorithm to the stored data to make a diagnosis concerning the vehicle.
  • the inspection system 1 further comprises a programmable logic controller Control (PLC) 6 for controlling the inspection system 1.
  • PLC 6 can have a PFC (programmable fieldbus controller).
  • the inspection system 1 comprises a lighting system 7 for illuminating an inspection section of the vehicle.
  • the lighting system 7 is connected to the PLC 6 for control purposes.
  • the inspection system 1 has a light barrier 8.
  • the light barrier 8 is also connected to the PLC 6 for control purposes.
  • the sensor assembly is used to inspect the inspection section 1. It is inventively provided that a technical functional state of the vehicle in the operating state, ie while the train is traveling, is inspected.
  • the first sensor module 3 includes three optical sensors 2a-c, namely area-scan cameras.
  • the cameras 2a-c are adapted to take black and white pictures.
  • the three cameras 2a-c of the first sensor module 3 are each arranged to record images with a data volume of about 4.5 MB per image.
  • the frame rate is around 5
  • a first camera 2a is connected to the PLC 6 via a first trigger line 9.
  • the first trigger line 9 is electrically grounded via a first ground connection 10.
  • the first trigger line 9 is designed as an optical cable, more precisely fiber optic cable, and connected to the first camera 2a via an optocoupler 1 1 for transmitting a trigger signal.
  • the first sensor module 3 further includes a first switch 12.
  • the first switch 12 has a module-internal connection, via which the first switch 12 can receive data from the optical sensors 2a-g, and a module-external connection, via which the first switch 12 data can transmit the server 5.
  • the three cameras 2a-c are each connected via a first combined Ethernet data-current line 13 to the first switch 12 at its module-internal connection.
  • the first switch 12 not only receives the data from the three cameras 2a-c via the module-internal connection, but also supplies each of the three cameras 2a-c via the module-internal connection, in each case with 48 volts of voltage and a maximum of 20 watts of power.
  • the data flow from each camera 2a-c to the first switch 12 is a maximum of 40 MB / s, whereby as a rule only actual values of about 22.5 MB / s are achieved (at 5 frames per second).
  • the first switch 12 sends to the server 5 according to data at up to about 125 MB / s (1 Gb / s) via a first switch server connection 14. This data in addition to the image data of the three cameras 2a-c one more certain administrative overhead.
  • the first switch 12 is over a second one Earthing connection 15 grounded.
  • the second ground connection 15 differs from the first ground connection 10. Since the first trigger line 9 is an optical cable, there is thus no electrical coupling between the first camera module 3 and the first one
  • Trigger line 9 The first trigger line 9 and the first sensor module 3 are, in other words, galvanically isolated. In this way, interference effects on the first sensor module 3 are reduced, which can have a positive effect on the recording quality of the three cameras 2a-c.
  • the second sensor module 4 includes four optical sensors 2d-g, which preferably have a distance of about 60 cm from the vehicle during application to the vehicle.
  • Two of the optical sensors 2d, 2e are line scanners with integrated camera, the other two optical sensors 2f, 2g are cameras without line scanners.
  • the second sensor module 4 is configured to perform laser line measurements, in particular for a light-slit method.
  • the raw data recorded by the cameras 2d-g of the second sensor module 4 are internally processed internally in the cameras 2d-g of the second sensor module 4 itself.
  • a raw height image and / or raw intensity image is generated in the second sensor module 4.
  • a calibration and rectification of the detected laser image is performed in the cameras 2d-g of the second sensor module 4.
  • the second sensor module 4 includes a second switch 16.
  • the four optical sensors 2d-g of the second sensor module 4 are connected to the second switch 16 via a second combined Ethernet data-power line 17, respectively. Since the data rate to the second switch 16 is reduced by the internal data processing in the cameras 2d-g of the second sensor module 4, to about 80 MBit / picture or a 6-channel picture to 12 Mbit per channel, it is unproblematic in this second sensor module 4 even four cameras 2d-2g via a combined Ethernet data-power line 17 to the second switch 16 to connect. For each camera 2d-2g less than 40 MB / s of image data is sent to the second switch 16.
  • the second switch 16 is connected to the server 5 via a second switch-server connection 18 and configured to send a maximum of 125 Mbytes / s (1 Gbit / s) to the server 5 at maximum.
  • the second switch 16 sends the captured raw data and the calibrated and rectified images to the server 5. Further evaluation is performed on the server 5 side based on the calibrated and rectified images.
  • One of the optical sensors 2d-g of the second sensor module 4 is connected to the SPS 6 by a second trigger line 19.
  • the PLC 6, the second sensor module 4 trigger.
  • the illumination system 7 is connected to the PLC 6 via a third trigger line 20.
  • the PLC 6, the lighting system 7 trigger.
  • the PLC 6 is connected to the server 5 bidirectionally via a PLC server connection 21.
  • the illumination system 7 is connected to the first ground connection 10, as is the first switch 12
  • Sensor module 3 with respect to the second sensor module 4 along the direction of movement of the vehicle spaced.
  • the first sensor module 3 and the second sensor module 4 are at the same height.
  • the server 5 preferably has eight Gbit network inputs for interfacing with the sensor array; seven for the sensor modules and one for the PLC 6, for example to provide updates.
  • the server 5 preferably has eight Gbit network inputs for interfacing with the sensor array; seven for the sensor modules and one for the PLC 6, for example to provide updates.
  • six sensor modules are present, which in each case resemble the first sensor module 3.
  • a preferred server 5 may manage 15-25 cameras, preferably 20 cameras, arranged as shown in groups in multiple sensor modules.
  • the trigger signals are sent by the PLC 6 via the first trigger line 9, the second trigger line 19 and the third trigger line 20 as soon as the light barrier 8 is triggered by a train passing through the inspection system 1 and the PLC 6 was woken up by the light barrier 8.
  • the illumination system 7 is activated and the first sensor module 3 and the second sensor module 4 of the sensor arrangement begin to record raw data.
  • the first sensor module 3 transmits the raw data unchanged to the server 5 for further processing.
  • the cameras 2d-g in the second sensor module 4 preprocess the raw data and send extracted data via the second switch 16 to the server 5. There, a further processing of some or all of the Sensors 2a-g recorded and received by the server 5 data.
  • the software resources and hardware resources provided in the inspection system 1 and in particular the data processing system thus permit
  • the software and hardware resources are for one
  • the server 5 is set up to carry a medical record via trains and diagnoses via inspection objects, such as
  • Server 5 may form an internal subsystem of the data processing system, while an external subsystem, such as e.g. a cloud computing farm, outsourced.
  • an external subsystem such as e.g. a cloud computing farm
  • the server 5 is set up to generate a three-dimensional model of the inspection section 1 from surface image data of the first sensor module 3. From several two-dimensional images of the moving train, which were recorded temporally sequentially by one of the three cameras 2a-c of the first sensor module 3, the three-dimensional model is derived by an algorithm on the server 5. For example, one of the three cameras 2a-c of the first sensor module 3 can be made five shots as the train passes.
  • a 3D model can be generated with only a single 2D camera or area-scan camera because the train moves in relation to the camera.
  • Figure 2 shows a schematic diagram of an embodiment of a
  • the inspection system 1 and the data processing system 5 shown in FIG. 1 are configured to carry out this inspection process.
  • a first step S21 the sensor arrangement is placed on the
  • Inspection section of the vehicle used to inspect the vehicle This is already known in general.
  • the sensor arrangement according to the invention the inspection section with an optical sensor 2a-g for inspecting a technical operating state of the vehicle applied S22, as described above.
  • an optical measurement technique a technical function of the vehicle in the operating state, ie in motion, inspected.
  • rail vehicles such optical inspections of the technical function in the operating state a large
  • the inspection method includes further steps. These result from incorporation of method steps from the general description of this application, from the detailed description and in particular from the
  • FIG. 3 shows a schematic diagram of substeps of an embodiment of an inspection method according to the invention.
  • the inspection system 1 shown in FIG. 1 and the data processing system, ie the server 5, are set up to carry out this inspection process.
  • a maintenance plan of all vehicles, in this case trains is read.
  • a pre-warning time for the inspection system 1 is set.
  • the inspection system 1 is initialized, in particular the sensor arrangement.
  • a functional test of the inspection system 1 is performed.
  • a functionality of the inspection system 1 is automatically determined.
  • it is selected depending on an inspection method based on historical data depending on the vehicle to be inspected.
  • environmental parameters are determined after applying the sensor arrangement.
  • steps S301 to S310 are executed in the order indicated. In other embodiments, however, a changed order may be used or further steps may be taken
  • FIG. 4 shows a further schematic diagram of substeps of an embodiment of an inspection method according to the invention. The substeps may be combined, for example, with the steps of FIG. 3, eg, entangled, or the steps of FIG
  • an inspection result for the vehicle is derived from a diagnosis made using raw data.
  • the inspection result is confirmed by a user. This can be done on a user client via a corresponding interface, e.g.
  • a third step S403 is by the
  • a maintenance action for the vehicle is proposed to the user. This maintenance action can be based on a medical record.
  • the vehicle is later inspected again by technical sensors, in this embodiment by the inspection system on a later passage of the vehicle, based on the previous diagnosis to confirm or reject the previous diagnosis. That is, the inspection method in this embodiment allows to check whether a diagnosis or prognosis actually becomes a reality or not.
  • a new diagnosis ie a previous diagnosis
  • a new diagnosis may be made after 4 weeks, if the re-diagnosis "pantograph is likely to break one week ", confirms the previous diagnosis. If the re-diagnosis is "pantograph breaks probably in 4 weeks” the previous diagnosis is discarded and a

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Traffic Control Systems (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
  • Geophysics And Detection Of Objects (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Automobile Manufacture Line, Endless Track Vehicle, Trailer (AREA)

Abstract

L'invention concerne d'une part un procédé d'inspection servant à l'inspection optique d'un véhicule à l'état de fonctionnement en utilisant un système d'inspection (1) qui comprend un ensemble de détection et fournit un passage pour le véhicule, pendant un déplacement relatif entre le véhicule et l'ensemble de détection, le véhicule étant de préférence un véhicule traversant le passage, lequel véhicule est, de manière particulièrement préférée, alimenté en puissance d'une ligne haute tension. Le procédé comprend l'étape consistant à utiliser l'ensemble de détection sur au moins une partie à inspecter du véhicule, afin d'inspecter le véhicule. Selon l'invention, le procédé comprend l'étape consistant à utiliser au moins un capteur optique (2a-g) de l'ensemble de détection sur la partie à inspecter pour inspecter un état de fonctionnement technique du véhicule. Ainsi, l'état de fonctionnement technique du véhicule peut être inspecté optiquement sans devoir arrêter le véhicule. Ceci permet de réduire les coûts et d'augmenter la durée d'utilisation du véhicule et présente de nouvelles possibilités pour les procédés d'inspection existants. L'invention concerne d'autre part un système de traitement de données qui est conçu pour mettre en œuvre le procédé d'inspection selon l'invention en liaison fonctionnelle avec le système d'inspection. L'invention concerne par ailleurs un système d'inspection qui est conçu pour mettre en œuvre le procédé d'inspection selon l'invention.
PCT/EP2017/078967 2016-11-10 2017-11-10 Procédé d'inspection, système de traitement de données et système d'inspection servant à inspecter un véhicule à l'état de fonctionnement WO2018087341A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/EP2018/080950 WO2019092248A1 (fr) 2017-11-10 2018-11-12 Procédé et système d'analyse de données brutes enregistrées avec un système d'inspection pour l'inspection optique d'un véhicule

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102016013329 2016-11-10
DE102016013329.5 2016-11-10

Publications (1)

Publication Number Publication Date
WO2018087341A1 true WO2018087341A1 (fr) 2018-05-17

Family

ID=60413180

Family Applications (5)

Application Number Title Priority Date Filing Date
PCT/EP2017/078964 WO2018087338A2 (fr) 2016-11-10 2017-11-10 Procédé de contrôle et système de contrôle d'un véhicule en état de fonctionnement
PCT/EP2017/078963 WO2018087337A1 (fr) 2016-11-10 2017-11-10 Module et système d'inspection pour inspecter des objets se déplaçant
PCT/EP2017/078967 WO2018087341A1 (fr) 2016-11-10 2017-11-10 Procédé d'inspection, système de traitement de données et système d'inspection servant à inspecter un véhicule à l'état de fonctionnement
PCT/EP2017/078966 WO2018087340A2 (fr) 2016-11-10 2017-11-10 Procédé de contrôle, système de contrôle ainsi que système et procédé de commande d'un moyen de transport
PCT/EP2017/078969 WO2018087343A1 (fr) 2016-11-10 2017-11-10 Procédé de commande d'un système de moyens de transport, système de traitement de données

Family Applications Before (2)

Application Number Title Priority Date Filing Date
PCT/EP2017/078964 WO2018087338A2 (fr) 2016-11-10 2017-11-10 Procédé de contrôle et système de contrôle d'un véhicule en état de fonctionnement
PCT/EP2017/078963 WO2018087337A1 (fr) 2016-11-10 2017-11-10 Module et système d'inspection pour inspecter des objets se déplaçant

Family Applications After (2)

Application Number Title Priority Date Filing Date
PCT/EP2017/078966 WO2018087340A2 (fr) 2016-11-10 2017-11-10 Procédé de contrôle, système de contrôle ainsi que système et procédé de commande d'un moyen de transport
PCT/EP2017/078969 WO2018087343A1 (fr) 2016-11-10 2017-11-10 Procédé de commande d'un système de moyens de transport, système de traitement de données

Country Status (1)

Country Link
WO (5) WO2018087338A2 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019092248A1 (fr) 2017-11-10 2019-05-16 Db Fernverkehr Ag Procédé et système d'analyse de données brutes enregistrées avec un système d'inspection pour l'inspection optique d'un véhicule
US11828728B2 (en) * 2016-12-05 2023-11-28 The Regents Of The University Of California Ultrasonic inspection of railroad tracks using railcar mounted ultrasonic receivers

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109216094B (zh) * 2018-08-27 2024-03-01 中国铁路设计集团有限公司 一种电气化铁路接触网上隔离开关的紧急操作装置
CN109807908B (zh) * 2019-03-01 2021-09-28 西京学院 一种用于真空管道检查维修的磁吸式步行机器人
CN110191554A (zh) * 2019-05-21 2019-08-30 安徽理工大学 一种基于plc渐变调光和机器视觉的路灯系统
DE102020202945A1 (de) 2020-03-06 2021-09-09 Gebhardt Fördertechnik GmbH Lagersystem und Verfahren zum Betreiben eines derartigen Lagersystems
CN112881437B (zh) * 2021-01-13 2021-11-26 华东交通大学 一种接触网探伤检测系统
CN113093595B (zh) * 2021-03-24 2023-03-10 陕西疆晨信息科技有限公司 电交换控制方法、装置及电交换控制箱
CN113238121A (zh) * 2021-05-20 2021-08-10 石家庄铁道大学 牵引接触网综合监测分析装置

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011035983A1 (fr) * 2009-09-28 2011-03-31 Siemens Aktiengesellschaft Procédé et système de contrôle de pantographes, de gabarits d'isolement et de la position horizontale et verticale du fil de contact au niveau des convois ferroviaires
EP2546120A1 (fr) * 2011-07-12 2013-01-16 ALSTOM Transport SA Procédé et système stationnaire pour surveiller l'équipement d'un véhicule de chemin de fer
DE102014226694A1 (de) * 2014-12-19 2015-12-31 Siemens Aktiengesellschaft Erfassung der Kontaktfehler zwischen Oberleitung undSchleifleisten

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE10240143B4 (de) 2002-08-30 2006-06-01 Siemens Ag Prüfung und Detektion potentialführender Teile und Leiterzüge mittels eines Foliensensors auf der Basis von Streukapazitätsmessungen
US6909514B2 (en) * 2002-11-18 2005-06-21 Beena Vision Systems, Inc. Wheel profile inspection apparatus and method
US9950722B2 (en) * 2003-01-06 2018-04-24 General Electric Company System and method for vehicle control
US8077480B2 (en) * 2008-11-26 2011-12-13 Rosemount Aerospace Inc. Faraday cage for camera
US9168937B2 (en) * 2012-10-24 2015-10-27 Progress Rail Services Corporation Multi-function dragger
WO2016033283A1 (fr) * 2014-08-27 2016-03-03 Lynxrail Corporation Système et procédé pour analyser des roues de matériel roulant

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011035983A1 (fr) * 2009-09-28 2011-03-31 Siemens Aktiengesellschaft Procédé et système de contrôle de pantographes, de gabarits d'isolement et de la position horizontale et verticale du fil de contact au niveau des convois ferroviaires
EP2546120A1 (fr) * 2011-07-12 2013-01-16 ALSTOM Transport SA Procédé et système stationnaire pour surveiller l'équipement d'un véhicule de chemin de fer
DE102014226694A1 (de) * 2014-12-19 2015-12-31 Siemens Aktiengesellschaft Erfassung der Kontaktfehler zwischen Oberleitung undSchleifleisten

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
PATRICK RAUL GIURGIU: "Support Vector Machine", Universität Siegen

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11828728B2 (en) * 2016-12-05 2023-11-28 The Regents Of The University Of California Ultrasonic inspection of railroad tracks using railcar mounted ultrasonic receivers
WO2019092248A1 (fr) 2017-11-10 2019-05-16 Db Fernverkehr Ag Procédé et système d'analyse de données brutes enregistrées avec un système d'inspection pour l'inspection optique d'un véhicule

Also Published As

Publication number Publication date
WO2018087340A4 (fr) 2018-12-27
WO2018087338A2 (fr) 2018-05-17
WO2018087338A3 (fr) 2018-08-09
WO2018087340A3 (fr) 2018-11-08
WO2018087340A2 (fr) 2018-05-17
WO2018087337A1 (fr) 2018-05-17
WO2018087343A1 (fr) 2018-05-17
WO2018087338A4 (fr) 2018-10-11

Similar Documents

Publication Publication Date Title
WO2018087341A1 (fr) Procédé d'inspection, système de traitement de données et système d'inspection servant à inspecter un véhicule à l'état de fonctionnement
WO2019092248A1 (fr) Procédé et système d'analyse de données brutes enregistrées avec un système d'inspection pour l'inspection optique d'un véhicule
US11854181B2 (en) Systems and methods for utilizing machine-assisted vehicle inspection to identify insurance buildup or fraud
DE102014105351B4 (de) Detektion von menschen aus mehreren ansichten unter verwendung einer teilumfassenden suche
DE102014106210B4 (de) Probabilistische Personennachführung unter Verwendung der Mehr- Ansichts-Vereinigung
EP3155603B1 (fr) Procédé de détermination de données de position à utiliser lors de l'exploitation d'un système de véhicule d'un véhicule automobile ainsi que système de détermination et de distribution de données de position
KR102017870B1 (ko) 실시간 선로 결함 검측 시스템
DE102016218012A1 (de) Verfahren für eine Datenverarbeitungsanlage zur Erhaltung eines Betriebszustands eines ersten autonomen Fahrzeugs sowie Verfahren für eine Datenverarbeitungsanlage zum Verwalten einer Mehrzahl autonomer Fahrzeuge
DE102020115550A1 (de) Fahrzeugwartung
DE102019002487A1 (de) Verfahren zur Überprüfung eines Umfelderfassungssensors eines Fahrzeugs und Verfahren zum Betrieb eines Fahrzeugs
EP3517398B1 (fr) Procédé de surveillance d'état de l'espace intérieur ainsi que véhicule doté d'un dispositif de surveillance d'état de l'espace intérieur
EP2521070A2 (fr) Procédé et système de détection d'une scène dynamique ou statique, de détermination d'événements bruts et de reconnaissance de surfaces libres dans une zone d'observation
WO2019092246A1 (fr) Procédé de contrôle et système de contrôle d'un véhicule en état de fonctionnement
WO2018095640A1 (fr) Procédé et système de détection d'un objet saillant se trouvant à l'intérieur d'un parc de stationnement
DE102016223094A1 (de) Verfahren und System zum Detektieren eines sich innerhalb eines Parkplatzes befindenden erhabenen Objekts
DE102020206128B4 (de) Verfahren zum Steuern einer flottenbasierten Zustandsüberwachung eines Straßenabschnitts eines Straßennetzes sowie zugehöriges System und Kraftfahrzeug und zugehörige Servereinrichtung
EP4157762A1 (fr) Procédé de détermination par machine de l'état de fonctionnement de rouleaux de support d'un système de transporteur à courroie, programme informatique et support de données lisible par machine
DE102020118589A1 (de) Verbesserter fahrzeugbetrieb
DE102019208448A1 (de) Verfahren zur Überprüfung eines Geräts oder einer Maschine
Di Summa et al. A review on deep learning techniques for railway infrastructure monitoring
Sacchi et al. PAVISYS: A computer vision system for the inspection of locomotive pantographs
DE102009021765A1 (de) Verfahren zum automatischen Erkennen einer Situationsänderung
Kumar et al. Improving Train Track Safety using Drones, Computer Vision and Machine Learning
Pradeep et al. Automatic railway detection and tracking inspecting system
EP3808687A1 (fr) Système de surveillance de l'état d'un convoyeur et convoyeur avec système de surveillance

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17801414

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17801414

Country of ref document: EP

Kind code of ref document: A1