MX2008000727A - Method, device and system for modeling a road network graph - Google Patents

Abstract

There is provided a method, device and system for modeling a road network graph, comprising the steps of receiving information data from a plurality of vehicles, said information data comprising at least positional data, and modeling said road networkgraph in accordance with said received data.

Description

METHOD, DEVICE AND SYSTEM FOR MODELING A GRAPH OF THE ROAD NETWORK FIELD OF THE INVENTION The present invention relates to the field of modeling (or generation or configuration or adaptation) of a road network graphic showing a simple topographic structure (shape, profile or contour, respectively) of roads, streets and other relevant connections from the point of view of traffic. In addition, a server device and a system are provided that have been adapted to effectively carry out a method for modeling said graph. BACKGROUND OF THE INVENTION As the number of vehicles has increased more or less constantly during the last two decades, at present and especially in developing and rapidly growing countries such as China, Russia and Brazil, there is also a demand for provide drivers with accurate navigation of roads and traffic, as well as providing road system planners with data that can help them cope with continually growing traffic. At present, some advanced systems have been implemented in the world to address both tasks. Developed nations have produced digitized models of Ref.188557 its road systems in the last couple of years, which allow drivers to find their way. At present, this system is known as vehicular navigation. These systems are sometimes complemented with systems that provide real-time traffic data that are usually purchased by a highway system operator and communicated to vehicles that are equipped with navigation devices that use a broadcast technology or similar (RDS, etc.). The totality of these systems requires a tedious collection and verification of data with geodetic means: (through the use of modern techniques such as GPS). On the other hand, the acquisition of data related to traffic conditions requires the installation of devices for recognizing deviations of the vehicle (called circuits, microwave curtains, cameras and the like), the communication of unusual events reported by the same drivers and the monitoring by vehicles, airplanes or helicopters, special. While the aforementioned measures are almost inevitable when it comes to real-time data, they provide less useful data to road system planners. BRIEF DESCRIPTION OF THE INVENTION The object of the present invention is to provide a methodology, a device and a system for modeling the graph of the road network, which overcomes the deficiencies of the state of the art. The objects of the present invention are achieved by the subject treated, defined in the appended independent claims. According to a first aspect of the invention, a method is provided for modeling a road network graph, preferably carried out on at least one modeling server. This method of modeling may include a method to calculate said graph, a method (preferably automatic) to outline said graph, a method (preferably automatic) to update, and a method to verify said graph. Said method comprises at least the following steps: receiving information from a plurality of vehicles, said information data comprising positional data, preferably geolocation data of said plurality of vehicles; and modeling said graph of the road network according to said received data. In this way an effective update of the graphic of the road network is achieved in a reliable and economical way. On the other hand, the method for calculating said graph may include an automatic calculation of the geometry of the road network (position data), of the topology (connection data), and of the statistics (amount of traffic, average speeds, etc.). In this way it is possible to obtain detailed traffic data and statistics, for use in navigation systems and in the apparatus for traffic control / planning. In addition, the method for calculating said graph can use measurements from the vehicles, said information being included in the system, or information from the network of the graph from other sources (government agencies, mapping companies or road builders, recognition of aerial photographs or other image formation, etc.). In this case, it is basically the fusion of graphics. It is therefore possible to merge information from various sources. On the other hand, the profiling method may comprise steps (preferably automatic) of abstract representation of roads and crossings and the establishment of their parameters, according to said information. For this reason it is possible to complete the graph and transform it into other graphic representations (more abstract). In addition, said information data can also be obtained from third parties, for example. This is especially the case of the type of information for which said plurality of vehicles (verification vehicles not included) are not equipped, for example street names or speed limits. In this way, another source is acquired. The method for updating said graph corresponds substantially to the method of its profiling. A basic difference consists in informing about significant changes in the graph and in a step of the calculation of graphs on the new sub-sections. The verification method may include the inspection of the graph, which is also carried out by specially equipped verification vehicles, which travel the road network and look for incompatibilities with said graph of the road network and which provide additional information about it. By using said graph of the road network, previously provided, an optimization step can be implemented within a certain process for verification. This optimization may consist in optimizing the roads for the verification vehicles. In this way another means is provided for cross-checking and for verification of the road network graphic. According to another embodiment of the present invention, said modeling is based on mathematical techniques for processing curves, arcs, polynomials or the like, carried out on said data. Therefore, said modeling can be implemented within a computer system, through the use of said mathematical techniques. That is, it is possible to process different data with the same method, for example, with which reproducible results are achieved, for example. According to another embodiment of the present invention, said modeling is based on Bezier curve techniques carried out on said data. It is preferable that it is possible to use Bezier curves due to the good approximations achieved in practical modalities through the use of said curves. According to another embodiment of the present invention, said information data may comprise the type of vehicle, the vehicle speed, the acceleration, and similar data. It is advantageous that said data may comprise additional information (mentioned above) that allows an improved modeling of said road graph. By means of said additional parameters it is easy to study certain behavior (driving) of a particular automobile, for example. According to another embodiment of the present invention, said received information data represents a trajectory of at least one vehicle among said plurality of vehicles. It is preferable that each path be described by Bezier curves. These Bezier curves allow an adequate and accurate representation of said trajectories, corresponding to the particular road of a particular vehicle. According to an advantageous embodiment, averaging the trajectories associated with said at least one vehicle can be provided. When averaging, it is possible to achieve an accurate representation of the trajectory. The main trajectory is calculated on the basis of a plurality of trajectories, resulting in an improved model. Said plurality may originate from a specific vehicle or even from different vehicles. According to another embodiment of the present invention, the calculation of a first approximation of said graph of the road network is provided on the basis of said received information data, the profiling of roads and crossings within said first approach results in a graphic of the road network, profiling, and carrying out a verification of said profiled network. The aforementioned steps improve the resulting representation of said graph of the road network. Said first approach is used as a first approach, and the subsequent steps can be carried out repetitively corresponding to a closed circuit, ie, said circuit corresponds to an advantageous implementation according to the present invention. According to another embodiment of the present invention, said calculation is based on Bezier curve techniques. It is preferable that said calculation can be based on Bezier curves that deliver exact and detailed results. According to another embodiment of the present invention, the detection of changes in a graph of the existing road network is provided on the basis of said information data received.; the storage of said changes; and the implementation of said changes in said graph of the existing road network. With this, the changes in an existing road network are detected and according to the present invention the methodology will implement the changes on the basis of for example a graph already modeled or calculated. According to another embodiment of the present invention, said implementation is based on statistical information. Generally said statistical information corresponds to information related to the transit such as the average speed of a vehicle traveling according to for example certain times of the week, time necessary to go from a place A to a place B by said vehicle or by using another type of vehicle, etc. Another transit condition such as said statistical information may be used. It is considered to still use the behavior of a determined driver as statistical information data. For example, a professional driver such as a taxi driver will have a behavior different from that of a normal driver who wants to go from point A to point B. In addition, it is considered that such statistical information is provided through a process of collection / obtaining by measurement vehicles or similar. According to another embodiment of the present invention, the transmission of the information related to said network graphic to at least one vehicle among said plurality of vehicles is provided. With which remote navigation of a vehicle can be provided. That is, a driver of said vehicle will receive navigation data from the server in such a way that it is possible to control the trip from a distance. It is advantageous if the actual information related to said graph of the modeled road network is transmitted to the driver to allow economic driving behavior, for example. Since the graphic of the road network adapts automatically and / or periodically, the driver of a vehicle will always receive real information about the characteristics of the road, for example. According to another embodiment of the present invention, the profiled graph of the road network includes information about the moments or times of transversal travel when the road is taken. It is advantageous if said times can also be used for navigation purposes, for example. It is also possible to implement road planning on the basis of said timing data. According to another embodiment of the present invention, said information from said plurality of vehicles can be compressed by the use of mathematical techniques to process curves, arcs, polynomials, etc. By means of said compression techniques it is possible to reduce the amount of data to be stored and / or processed. According to an advantageous embodiment, the trajectories can be described by Bezier curves. According to another embodiment of the present invention, the step of compressing said information data selectively within said modeling entity and / or within said plurality of vehicles is provided. It is therefore possible to perform compression on the vehicle side, which means that the server entity can be released, that is, it is possible to use the electrical energy saved from computing, for other purposes. In accordance with another embodiment of the present invention, storage of said information data is provided. In this way the future use of certain data of interest is ensured.

According to another embodiment of the present invention, said calculations are based on digital computing techniques for the exact calculation of the values of fixed points. In this way it is possible to provide such calculations to the entities based on fixed point architectures. According to another embodiment of the present invention, said information data comprise measurement data, and further a step is provided for the normalization of said measurement data according to predetermined threshold values. By carrying out said step of normalization, the data will be represented according to predefined thresholds, which improves the handling and / or illustration, for example. It is also possible to apply it in entities based on fixed point architectures, and with which the calculation error is reduced. According to another embodiment of the present invention, said storage is provided after execution of a compression algorithm, a key generation algorithm (coding with keys), an encoding algorithm or the like. In this way a secure and compressed storage of the data is achieved. Therefore, each record is sent after the on-board device has determined all of the necessary information. According to another embodiment of the present invention, the detection of the existence of a phenomenon / effect of multiple trajectories is provided, and in this case it is possible to assign a lower weight to said information received during said calculation step. In this way it is ensured that the data that has been corrupted due to the effect of multiple trajectories, will receive a lower weight during the steps of the calculation, for example. According to another embodiment of the present invention, the measurement of the dimensions of the roads is provided by a provider of position information, within said plurality of vehicles. In this way, a characterization of the road axis is provided, corresponding to the shape of the trajectory. On the other hand, a detailed sizing of said road axis corresponding to the width of the street (road), etc. is provided. According to another embodiment of the present invention, said entity is a GPS transceiver located within said vehicle. However, said transmitter-receiver is adapted to receive and / or send positional data to a suitably equipped vehicle. According to another embodiment of the present invention, the calculation of the geometry, topology and statistics of the road network is provided in an automatic manner. These calculations are carried out automatically by means of a periodic algorithm, for example. According to another embodiment of the present invention, automatic profiling of the road network graphic is allowed by the use of said information data. According to another embodiment of the present invention, an automatic update of the road network graphic is allowed, also by the use of said information data. Said automatic profiling and / or updating can also be based on periodic algorithms, for example, which are repeated on a time basis. In accordance with another aspect of the present invention, a computer program product is provided, comprising program code sections stored in a machine-readable medium for carrying out the operations of the method in accordance with any of the aforementioned embodiments of the invention. invention, when the computer program product is executed in a device based on a processor, a computer, a terminal, a network device, a mobile terminal, or a mobile communication enabled terminal. According to another aspect of the present invention, a computer program product is provided, comprising program code sections stored in a machine-readable medium for carrying out the operations of the aforementioned method according to a mode of the invention. present invention, when the computer program product is executed in a device based on a processor, a computer, a terminal, a network device, a mobile terminal, or a mobile communication enabled terminal. In accordance with another aspect of the present invention, a software tool is provided. The software tool comprises program parts to carry out the operations of the aforementioned methods, when the software tool is implemented in a computer program and / or executed. According to another aspect of the present invention, a computer data signal incorporated in a carrier wave and representing instructions is provided, said instructions, when executed by a processor, cause the operations of the method of operation to be carried out. according to a previously mentioned embodiment of the invention. According to yet another aspect of the present invention, a server device is provided for modeling a road network graphic. Said server device comprises at least one component for receiving information data from a plurality of vehicles, said information data, position data and a component for modeling said graph of the road network according to said received data. According to another aspect of the present invention, said server also comprises a component for calculating a first approximation of said graph of the road network; a component to profile roads and junctions within said first approach, the result being a profiled graph of the road network; and a component for carrying out a verification of said profiled network. In this way, all the elements located within said profiled grid chart are updated, based on the data that originated from said plurality of vehicles. This means that all the elements will receive additional attributes based on vehicle data. Said operation of profiling can also be provided periodically to ensure a constant updating of said elements of the network. Additionally, some of the attributes that can be used for profiling operation can be gathered from other databases such as government databases, road construction companies, etc. The data corresponding to said attributes can be inserted manually and / or automatically for further use within said profiling step (and also modeling).

It should be noted that all of the information collected can be stored and used at any time. According to yet another embodiment of the present invention, said server further comprises a component for detecting changes in said road network graphic on the basis of said received information; a component to evaluate those changes; and a component to include said changes in said graph of the road network. According to yet another embodiment, said server further comprises a component for analyzing said graph of the road network on the basis of said received information; and a component to report the results of the analysis to a third party. According to yet another embodiment of the present invention, said server further comprises a component for performing a compression step of said information, selectively within said modeling entity and / or within said plurality of vehicles. According to yet another embodiment of the present invention, said server further comprises a component for storing said information. According to yet another embodiment of the present invention, said server further comprises a component for detecting the existence of a phenomenon / effect of multiple trajectories; and also a component to assign a lower weight to said received information. According to yet another embodiment of the present invention, said server further comprises a component for measuring road dimensions by a position information providing entity located within said plurality of vehicles. According to yet another embodiment of the present invention, said received information represents a trajectory of at least one vehicle among said plurality of vehicles, wherein each path is described by Bezier curves, for example, and said server further comprises a component to average trajectories associated with said at least one vehicle. According to yet another aspect of the invention, a system is provided for modeling a graph of the road network, said system comprising a plurality of serving devices and a plurality of vehicles providing information data. Furthermore, according to a preferred embodiment of the present invention, Bezier curves can be used to model said road network graph. 1. Throughout the detailed description and the accompanying drawings, the same or similar components, units and devices will be referred to by similar reference numbers, for reasons of clarity.

BRIEF DESCRIPTION OF THE FIGURES The attached figures are included to provide a better understanding of the invention, and said figures are incorporated in this descriptive specification, of which they are a part. The figures illustrate embodiments of the present invention and together with the specification serve to explain the principles of the invention. In the figures: Figure 1 shows a flow chart illustrating the principle of the method according to the present invention; Figure 2A shows an operational sequence according to the present invention; Figure 2B is a flow diagram showing the principle of change detection according to the present invention; Figure 2C shows an analysis and representation of real-time reports of traffic data according to the present invention; Figure 3 shows the principle of a system according to the present invention; Figure 4 is an on-board unit device according to an embodiment of the present invention; and Figure 5 is the principle of an automatic registration according to another embodiment of the invention; Figure 6 shows the principle for averaging various trajectories, represented by Bezier curves. Although the invention has been described above with reference to modalities in accordance with the appended figures, it is evident that the invention is not limited to this, but it is possible to modify it in various ways without departing from the scope of the appended claims. DETAILED DESCRIPTION OF THE INVENTION The following description introduces a system according to the present invention, which provides a generation and verification of a digital model, preferably vectorized (or described with curves) of the road network, the efficient updating of the digital model of the road network, the profiling (.configuration of attributes) of the digital road network. To achieve the aforementioned task, the system uses stored road data received from a large number of vehicles equipped with position receivers (GPS, GALILEO or similar) that transmit their position and other data to a server. These receivers are preferably also equipped with wireless data transmitters, which transmit the data stored on roads traveled at certain times, with more or less frequency, in which according to another option, the data from the receiver will be read manually and subsequently transferred to a central storage. The amount of data, which describes the roads accurately, is huge. This is why a special compression is needed - either before transmitting the data to a local exchange (to reduce the cost of communication) or before storing it. All data can be stored on a remote server or in a multitude of them, and special software tools can be used to combine all the data available for a desired set of data corresponding to a geographical area to be analyzed. Figure 1 schematically shows the principle of the present invention on the basis of a data flow diagram. The operational sequence according to the invention can be initiated by any method. Said start operation can be provided automatically, by means of the capture of the user, or similar. It is considered that the operational sequence will be activated or initiated, respectively, if new data is received or determined. In the next operative step 100, the reception of data is provided, wherein said reception of the data can be a process that is repeated continuously or periodically. This operation corresponds to the acquisition of the data, which is described below. In a next operative step, this graph of the road network is modeled at 150. All calculations and modeling operations can be based on Bezier curves as described below. Once all the modeling and calculation steps have been completed, the methodology can reach an END, and can be restarted, which corresponds to a new operation according to Figure 1. Also, the modeling step is considered 150, you can receive additional information from other entities located within the system. This means that new or similar repetitions can be controlled by means of external processes or operations, or even by capturing the user, for example. While additional parameters corresponding to information from said plurality of vehicles are received, it is possible to restart the modeling step 150, until a desired result is achieved. Referring to Figures 2A to 2C, the system may operate as follows. In general there are three basic processes. The first process, Figure 2A, is an initial calculation that gives the first result of the graph of the road network. The second process, Figure 2B, may be repeated periodically, for example once a month. It provides the system with a regular update of the changes in the road network system. These changes can either correspond to changes in the magnitude of the road network (geometry and / or topology), or in terms of its statistics (attributes). Within the scope of the present invention it is possible to implement other changes that have to be used for issues related to the update. The third process, Figure 2C, is constantly analyzing the current traffic situation. If a special situation is detected (with a high statistical probability), the system informs the appropriate recipient (traffic control center, police, etc.) about this. Referring to Figure 2A, the process representing an initial calculation of the road network graph is illustrated. In a first operational step, a data collection 200 is provided. This means that a plurality of suitably equipped vehicles deliver / send position information to a central server, for example. It is considered that said shipment is provided periodically or even manually. This means that the data obtained, currently located in a storage of said vehicle, must be transmitted in some way to said server or central provider, for example. In a next operational step 210, it is possible to provide the calculation of a first approximation of said road network, said approximation corresponding to an initial graph of the road network. According to the first set of position information, it is possible to carry out a first calculation of an approximation of the graph. This first approach will correspond to a provisional representation of the road network, and of course it must be corrected or revised. The profiling of roads and / or crossings 310 is then provided. During this step it is possible to provide some parameters such as the direction of the road and / or type of crossing, as well as other attributes such as average speed, for example, time (necessary to travel a certain connection or distance, respectively) or similar, which can be done in accordance with step 215 verified. Verification step 215 can provide a first verification of the first approximation and subsequently said graph can be constantly reinforced and / or expanded. The main difference between steps 215 and 225 is that step 215 is preferably carried out on the entire graph while step 225 is carried out only on certain detected / determined changes. With reference to Figure 2B, the principle of updating and updating said first approach is illustrated. The step of data collection is similar to the aforementioned step according to Figure 2A. The vehicles, properly equipped, constantly deliver position information, among other data. Said data may also comprise information about the type of vehicle, driver, etc. In the next step 220, a comparison can be provided between the existing data, included in the existing graph, and the newly received data. As a result, it is possible to signal a list of changes or even new roads, etc., so that the methodology may be able to update this first approach. Said step of the update is illustrated with reference to operational step 225 in Figure 2B, and such changes may comprise changes to the structures of the chart such as for example the omission of existing roads or the addition of new roads or even of their attributes (for example, speed, time, traffic rules, etc.). It should be noted that the input to step 220 (Figure 2B) can be either the result of the operational sequence according to Figure 2A (or of some other graphic), or, in the future, the output of the sequence of agreement with Figure 2B and additionally of Figure 2C. Figure 2C shows an operational sequence according to the present invention, in which a real-time analysis of the traffic conditions is provided and subject to a subsequent report. As already mentioned above, a data collection 200 is continuously provided and the system according to the present invention is capable of analyzing the existing traffic data. This analysis, 230, can be based on probability theories in such a way that it is possible to face a probabilistic and / or predictive operation of traffic monitoring. According to the present invention, the results of said analysis, 230, may be the subject of a subsequent report to third parties. Said third parties may correspond to a central traffic monitoring institute or even to a vehicle or driver, respectively. Within the scope of the present invention there are many configurations. The object of the acquisition or collection of data is set out in detail below. The device located in a vehicle (on-board device) among said plurality of vehicles, can provide its position for which it uses for example a GPS signal (it could also be another similar system, such as Galileo), and eventually some devices for the approximate calculation of the position (for example a gyroscope) every second, since it is usually the smallest time interval that the GPS receivers can attend. If the measurements were connected to each other by straight lines, they could describe the shape of the road very well. The problem that is achieved is the amount of this data. This is why a compression is needed. If the amount of data were reduced, few advantages would be achieved: the reduction of data transfer to the central server, the decrease in the size of the database, the (post) processing time could be reduced. Also, it is considered that the shape of the road is described with great precision, so that the error does not exceed the width of the road or, in general terms, the geometry of the road. Therefore, adequate compression, substantially without loss, of the form is needed. To this end, third-order Bezier curves can be used to describe the shape of the road. The Bezier curves are a very flexible and geometrically simple representation. These curves can describe "U" and "S" shapes, closed curves and circuits. Other curves could also be used, similar to Bezier curves, of higher order such as arcs, polynomials, etc. Another characteristic considered would be to also describe other information data, not just the shape of the trajectory. Along with this it is possible to describe and make available data such as speed, rpm of an engine, etc. In general terms, the term "trajectory" refers to the journey or journey of a vehicle in a given environment. This means, according to a trivial description, that the path of a specific car can be represented by a line (curve), in which each point of that line describes the actual geographical position (it is also possible to include the altitude), of the vehicle . It is also considered that each point of the trajectory will be associated with the actual speed, acceleration or the like of the vehicle, which is advantageous for later calculation or modeling subjects. Setting the Time Interval between two Records The time interval between two records depends considerably on the shape of the road. The word "register" refers to the storage of certain information from said plurality of vehicles. The on-board device can record various positional data before sending it to the server. Such positional data correspond to a road trip (trajectory) of said vehicle. Data can be sent spontaneously without storage, or as already mentioned, positional data can be accumulated (main purpose of component 415) and sent later. Generally speaking, a large part of a road can be approached well by a simple curve; while on the other hand, a winding mountain road has only a brief part of it that can be described with a single curve. The time interval is usually longer on main roads. The object consists in obtaining a description of the road (the trajectory of the vehicle) with a minimum number of elements and also a minimum of errors. For this a heuristic approach may be necessary. The on-board device has a buffer memory, which contains a series of consecutive measurements. The length of the buffer is equal to the length of the longest time interval between consecutive records (if the measurements have valid positions - if the on-board device is not without a gyroscope in a tunnel or in a garage). It is advantageous that it is possible to configure the shortest time interval. In this way it is possible to achieve a lower limit and an upper limit of the quality of the compression, according to the invention. On the other hand, since all of the measurement data is not available (said buffer is too small), a heuristic approach can be used to determine the appropriate representation of a given vehicle trajectory. The basic idea is that the measurements in the buffer memory are approximated by a curve (for example, a Bezier curve) at predetermined time intervals such as every second, for example. If the approximation already achieved is sufficiently good, some of the measurements may be omitted so as not to exhaust the resources to calculate the approximation in the future. If the approach exceeds a predefined error threshold, the process must stop and record (store) the existing curve with the measurement at the end of it and empty the buffer. This is how it is possible to ensure a small error (less than a predefined threshold) (without considering the GPS error) in the road description. There are also other conditions that trigger the recording of current measurements. According to the present invention, those measurements having a large derivative of the speed, preferably greater than a second derivative of the speed, used as a reference can be recorded. At these points the acceleration changes very abruptly. The shape of the road changes gradually if the acceleration is constant. It is easier to describe the shape of the road between the points of maximum second derivative of the speed. In accordance with the present invention it is possible to configure a threshold for the second derivative. If this threshold is exceeded in a given measurement, then it is possible to record a curve corresponding to that measurement (together with it). In this way a minimum amount of elements is achieved in the description of the road, according to the present invention. The current (or most satisfactory) curve and the measurement are recorded, if an abnormal behavior of the GPS signal is found, such as a phenomenon of multiple trajectories or loss of signal (when in a tunnel). In this way it is possible to avoid errors of erroneous measurements. A phenomenon or effect of multiple trajectories means respectively that the GPS signals from the satellites are reflected or can interfere with other signals, so that the data or signal communication may be erroneous. In this case, the receiver incorrectly determines the current position. The basic idea mentioned above can be applied to other quantities (for example, speed), and not only to the shape of the road. The measurements of this amount are approximated every second, and if the approximation is not good enough, it is possible to stop the approach process, and it is also possible to record the last satisfactory approximation. If increasing quantities (numbers) are observed, the use of polynomials instead of curves is considered, for example. Experimental observations show that said aforementioned approach allows recording every 30-40 seconds (on average) while accurately describing the shape of the roads with a tolerance of a few meters. These observations were approximated by third-order Bezier curves according to the present invention. Otherwise, the time difference can vary substantially. In general, the greater the order of the curve, the greater the difference in time between the records (time between two successive records of position). Effect or Phenomenon of Multiple Trajectories One of the most important problems when it comes to accurately describing the shape of a road, is the phenomenon of multiple trajectories. If it lasts a short time interval, it can be detected from the coincidence of: the difference in the address, informed by the GPS receiver, and the address, calculated from the GPS coordinates, and a greater estimated error of the coordinates. If this phenomenon is detected, then the measurements that intervene in it are assigned lower weightings than others when such measurements are approximated, according to the present invention. Therefore, the most accurate measurements have a greater influence on the shape of the curve.

It is considered that the measurements (and curves) are recorded or stored before the phenomenon takes place. This is because the measurements (or curves) of before the phenomenon, are not corrupted. If the phenomenon does not exceed the maximum time interval, it is preferred not to register anything until the phenomenon ends. However, the correct curves or approximations are based on the correct measurements considerably. If a multi-trajectory effect was determined, it is considered that the measurements taken within this period (during the effect of multiple trajectories) are not taken into account. The same also applies if only the estimated error increases. Application of the Kalman Filter within GPS Devices Another difficulty arises because a filter of Kalman in a GPS receiver, as is known in the art, does not work perfectly if the speed of the GPS receiver is low. Therefore, the location reported by GPS suffers from a slip every time the vehicle stops. This can be a serious problem in urban areas where there is a plurality of traffic congestion. The solution to this problem is to not register anything if the speed of the vehicle is low (for example, less than 3 km / h). In accordance with the present invention, the measurement (with the curve) can be recorded as soon as it is detected that the vehicle has stopped and immediately after it starts to move. The measurement made with low speed can be discarded, and in addition any approach steps are inhibited, and only consecutive records are connected (just before the vehicle stops and just after it starts to move) with a right (linear) curve . The two problems described above are solved if the on-board device has a device for approximate position calculation (a gyroscope), but this increases the price of the on-board device. Another problem is the limit conditions: the treatment of the start and end of the operation, temporary failure of operation, etc. In accordance with a possible embodiment of the present invention, it is possible to perform an implementation as follows. Therefore, the following quantities are under observation every second: - GPS coordinates - Position (P (t)), Estimation of the horizontal error (Sigma), calculated by the GPS receiver, Velocity vector, calculated by the GPS receiver (Azimut WGS84, Velocity (knots)), - Vector speed, calculated from the GPS coordinates ((P (t + 1) -P (tl) / 2), Acceleration (from the GPS heading), Acceleration (from the GPS coordinates), Derived from the acceleration (from the GPS heading), Derived from the acceleration (from the GPS coordinates), - Information about the validity of the data (see following enumeration): • 0 without heading, without coordinates, "1 without heading, coordinates OK, • 2 heading OK, without coordinates," 3 heading OK, coordinates OK. The preceding enumeration is given only by way of example, and the present invention is not limited thereto. It is also necessary to know if the position has been calculated by the GPS receiver or by means of an approximate calculation device of the position. Additionally, within the scope of the present invention it is also possible to observe other quantities. If the velocity vector, calculated by GPS receiver (Vs), and the velocity vector, calculated from the GPS coordinates (Vk), differ considerably from each other, and the error estimate (sigma) increases, a cause very likely it can be the phenomenon of multiple trajectories. A series of these measurements is stored in a buffer. The length of this buffer (Max) is the maximum time interval for an approximate curve. For a curve of this type, a minimum time interval (min) can be set. However, said interval provides a lower limit for the quality of compression and allows not recording the last measurement in the buffer memory. Also, it is possible to record a measurement that had been obtained up to a few seconds before the current measurement. If a measurement that was obtained r (<; min) seconds before the current measurement, then the buffer is not completely emptied - the last r measurements may be inside the buffer. If a circular buffer is used, it is not necessary to move these measurements to the beginning of the buffer. Therefore, the implementation according to an embodiment of the present invention allows storing the initial position and the current position in the buffer. Sometimes the record of a measurement is considered before the current measurement. Sometimes several consecutive measurements are needed to discover a certain phenomenon. For example, it is possible to use five consecutive measurements to calculate the derivative of the acceleration in the average measurement (the third measurement). In the current second, the derivative of two seconds ago is calculated. If said derivative is sufficiently large, the measurement (together with the curve) two seconds ago can be recorded according to the present invention. The buffer is then emptied, and only the last three measurements (= r) remain in the buffer. The unit should not perform any approximation for a few seconds, until there are min. Measurements in the buffer. From then on, the usual routine continues. The derivation function is smoothed by the use of orthogonal polynomials over five consecutive measurements. An additional buffer can be used, which stores the last few approximate curves, if for example the need to register a curve a few seconds ago is desired. In general, there are some limit conditions. The first measurement (with valid position) has to be recorded. The same applies to the last position, once the engine has been turned off. The last position outside a tunnel (with valid GPS position) has to be registered. It is also considered to set a threshold u, for how many consecutive seconds the GPS position must be invalid in order to mark it as the start of the tunnel. The object is to discard very short tunnels or errors, noise in GPS receivers. Once a measurement has been recorded as the start of a tunnel, the first measurement with a valid GPS position as the end of the tunnel must be recorded. If the on-board device does not have a device for the approximate calculation of the position, these two registers are connected to each other by means of a right curve, which is a line. The time interval between the two registers can be more than Max, in this case only. If the on-board device does not have a device for the approximate calculation of the position (a gyroscope), the registration procedure inside the tunnel is the same as usual. In the following section, the choice of the log step according to one embodiment of the present invention is described. For example, the following three quantities (values) are observed at a given moment, t: A (t) = magnitude of the derivative of acceleration (scalar), V (t) = difference of velocity vectors | IVs-Vkl | (scalar representation), S (t) = Sigma, estimated error (increasing value). If A (t) exceeds a predefined threshold, then the measurement is a member (subject) for the record. If a weighted sum of V (t) and S (t) exceeds another threshold (due to the possible presentation of the effect of multiple trajectories), then: • If (t-1) > min, then the preceding measurement must be recorded (so as not to corrupt the current approximation of the curve); otherwise • If t < Max, should not be recorded in the t-th measurement (since the effect of the multiple trajectories can be determined quickly, so that the correct termination of measurements and approximations of curves is possible). We want to find and record a measurement with a large derivative and small estimates of multiple trajectories and errors. There may be two limit values: minimum time for the new record (min), maximum time for the new record (Max). With reference to Figure 5, an automation is provided in accordance with the present invention. For example: (L represents LOG, 530, m, 510 is a measurement every second). This is an automation (according to Figure 5) that has a temporary state: Lmm? Rüi? Rrmtmmmmmmmmmmmmm ... = L + c * m This automation carries out a basic circuit: L + c * m If the quantity of Current measurements c, is greater than min and less than Max then if the trigger is set to t, t >; min, t < Max, (c-t) < min, the measurement t in the series is recorded, and the series is emptied to L + (ct) * m otherwise L + (c + l) * m goto Loop The trigger (520) may consist of several parts: A) Yes the second derivative of the velocity is higher than the pre-established threshold, this means that the measurement at tl: = c-2 is a candidate for the record; B) if the multipath phenomenon has probably taken place at t2: = c - 1 (the difference between the addresses is large and the estimated error has increased), then 1. if m (t2-l) does not have multiple trajectories and (t2-l) > min, then m (t2-l) is a candidate for registration, otherwise 2. if t2 < Max, m (t2) should not be registered; C) If the curve calculated in c does not agree sufficiently with the measurements, and the curve in t3: c-1, if it does, then m (t3) is a candidate for a record; D) if other increasing quantities are observed (speed, for example), and the approximation of the measurements is not good enough, then m (t4) together with the curve and the approximation function of this quantity should be recorded, t4: c -1. Then the minimum tm of candidates to register is selected (ti, t2, t3, t4). The new record is m (tm) with the corresponding curve and possibly approximation functions of other quantities. When trying to match a curve with position measurements, weights are weighted with the weight that decreases with increasing probability of multiple trajectories. If the agreement is made in fixed point arithmetic, some special measures must be adopted. Other limit conditions are also involved that must be taken into account: the first valid position is registered after the start; the last valid position is recorded (when the engine of a car is turned off); the last measurement is recorded before a tunnel (before the GPS positions lose their validity); The first measurement is recorded after a tunnel. Bezier curves A brief introduction to third-order Bezier curves follows, where advantageous adaptations are provided in accordance with the present invention. In general terms, these curves are defined by four control points, PO to P3. The curve is located inside the convex hull of the control points. The curve starts at the first control point and ends at the last. The direction of beginning of the curve is equal to the direction between the first two points, and said determination is equal to the direction between the last two points. From the numerical point of view, the curves of Bezier are defined by Bernstein polynomials on the control points Pk. B (t) = ypk M tk () N-k, - 0 = t = 1 describes the Kí (N-k) \ curve, parameterized by t. These curves can be divided by the De Casteljau algorithm (not shown). Another possibility is to match the Bezier curves with the measurements received or provided. If the mobile units (or devices) have a digital fixed signal processing unit, only fixed point arithmetic can be used, so the computational error due to the computation of fixed points has to be minimized or avoided. A first improvement according to the present was to include CORDIC algorithms (digital coordinate calculation) to calculate vector (or curve) norms, etc. The second improvement according to the present invention consists in selecting a limiting box (not hermetic) of measurements and normalizing them according to the magnitude of the limitation box and number ranges (arithmetic of fixed points). The current state of the art teaches adjusting only the length of the tangent (control) vectors of the curve (between the first pair and the second pair of control points), but it is also necessary to modify the direction. The following shows to what extent a greater flexibility in the shape of the concordance curve is achieved, according to the invention. The following definitions are established: Vi = Vo + ai i + ß? T? P V2 = V3 + a2t2 + ß2t2p where Vi are control points of the curve, and ti are control vectors (tangent) at the ends of the curve. curve, tjP is perpendicular to tj. OÍJ represents the correction of the length of the control vector; ßj represents the correction of the address. The solution for the values of ßj is similar to the solution for OÍJ, which is described in the prior art. The matching procedure can be repeated in a circuit, and the circuit can comprise two steps: first, adjust the length and second, adjust the direction of the control vectors.

In accordance with the present invention distance measurement can be provided by the use of GPS or information signals, respectively. It is possible to measure the length of a road with the help of the GPS system. If there are measurements available, they are taken every second (some may be missing), it is considered to add the distances between all the consecutive pairs and obtain a very accurate estimate of the real length. If the speed is very low (for example, less than 3 km / h), according to one embodiment of the present invention, the measurements can be discarded. Data Storage All the data and information used in the present invention and received from a plurality of vehicles, can be stored in a central location (server), and can be analyzed later in a couple of stages, for example to achieve the desired result. It is preferable that these data entered receive the denomination of raw data. The raw data may include at least one of the following: position, speed, bearing (address), moment of data acquisition, but may also include the following: a description of the curve (trajectory), a description of the function of other quantities (speed, etc.), an estimate of the horizontal accuracy of the position received by the position receiver, the number of satellites (GPS) with a good signal, the data from other sensors of the vehicles (temperature, weight), etc. The raw data can be stored in such a way that the trip (path or trajectory) of a vehicle is stored as a separate set of data, but nevertheless the identifier of the vehicle can be coded (coded with keys) or not yet present, to maintain privacy. The vehicle data can comprise two attributes to further help in the identification of road data: type of vehicle (passenger car, van, truck, bus, motorcycle, construction vehicle, tractor, ...), type of service (passenger, police, works, taxi, municipal truck, military use, agricultural use, ...). These two attributes mentioned above, can help differentiate the public network of roads and roads used by special types of vehicles (such as tractors) and roads used by private services with extended or limited rights (police, military, taxi, etc. .). Computation of the road network We start by analyzing the raw data to provide vectors (curves) that represent roads, organizing them in the form of a directed graph (as in the theory of graphs, well known in mathematics). This process requires a small number of very accurate measurements (as in the traditional geodetic approach) or a large number of less accurate measurements, which produce high accuracy, when averaged. According to the present invention, the focus is configured on the second situation. The edges of the graphs are the streets, and the vertices of the graphs appear when several roads are connected to each other. The vertices geometrically closest to each other represent the crosses. Therefore all the following operations are derived from the theory of standard graphics. The resulting graph is the graph of the basic road network. Simply put, the analysis transforms raw data from many vehicles that have traveled in the same way, in a vector (curve) that represents the road traveled. This process is not completely trivial. It is considered to keep in mind that the data may not accurately represent the traffic rules since it is possible that some drivers commit traffic infractions. The first objective is to produce a 2D map. It is also possible to include information about the height above sea level, if the measurements are sufficiently accurate. It is necessary to properly calculate two properties of the road network: geometry, which means exact positions of the axes of the roads, the topology, which means correct connections between the roads. The geometry is computed basically by the average of the trajectories of the vehicles that were on the same road. The topology is basically calculated by verifying which trajectories connect with the roads. There are several strategies to calculate the road map. Two basic approaches are described: a local version and a global version. In both versions it is possible to define the distance between the sampled points of the roads. The local version has a more local focus (in terms of distance). Progress locally by pre-established distances between sampled points. It is centered on the density of the resulting graph. This calculation of the map is based on two steps: the calculation of the sections of the roads and the calculation of the crossings of the roads. The basic operation consists of calculating a simple curve between sampled points, corresponding to an average of the measurements. According to experimental tests, a distance of 100 meters was chosen between two sampling points. According to the present invention, it is preferred to describe sections of a road between the sampling points as a straight line if the distance between the points is approximately 20 meters. Thus, the error produced is not significant, and the section of the road is adequately represented. In accordance with the present invention, Bezier curves can be used to represent vehicle trajectories and their calculated averages in the graph, due to their numerical stability and geometric flexibility and clarity. Average of the Bezier curves This procedure is part of the present invention and is used to calculate the geometry of the roads, but it could also be used for other purposes. According to the present, an initial observation of a plurality of trajectories provided by a plurality of vehicles that perform measurements is provided. Each of the trajectories of each vehicle is described by consecutive Bezier curves, according to the present invention. These curves usually have different lengths. In order to obtain the exact geometry of the road axis or of the sub-sections of the road, respectively, an average step of all the trajectories present can be provided, according to the present invention. The averaged curves must be sufficiently short to describe all the details of the road network with sufficient precision. Therefore, averaged Bezier curves having a length less than 100 were employed. The following section describes the step of the average of a set of trajectories described by Bezier curves according to the present invention. One of the objects is to average several trajectories. First, a starting point and a completion point can be chosen for each average. The starting point and the termination point between which the trajectories are averaged, can also be configured as a line, which is perpendicular to the trajectories, according to the invention. It is assumed that the average path between the starting point and the termination point (line) can be described with sufficient accuracy by a Bezier curve according to the invention. Before averaging, it is possible to divide the curves given in the points closest to the starting and ending points (or lines) chosen. Therefore, a result is obtained according to the sub-sections of the trajectories, which are very similar. There may be various ways to carry out said average, in accordance with the invention: 1. If all of the sub-sections between the starting point and the completion point are described by simple curves, it is preferred to simply average the points of control of the sub-sections. Otherwise you can choose another way to average, which is described below. 2. Average: "start and end coordinates in this sub-section" the velocities (lengths of the control vectors) in these start and end coordinates "lengths of the sub-sections" difference of the time of each sub-section section between the start and end coordinates. Therefore, sufficient data is provided to guess the trajectory (see next section). 3. Match a new Bezier curve with the measured positions (coordinates - points in the original curves) in this sub-section (see section on data compression). If there are not enough measured positions, it is preferred to add points arbitrarily over the curves. It should be mentioned that you do not have to use positions that could have a big mistake. The use of positions that are very close to the starting point or the ending point in each sub-section, can be unfavorable, since the error of these positions has more influence on the shape of the curve - so sometimes Small circuits may appear. Guessing or determining the trajectory In the case where trajectories with Bezier curves are not described, but the measurements are sufficiently accurate, it is possible to guess the trajectory and describe it by guessing a Bezier curve, as follows. Without compression, the data coming from the vehicles consist of positions, directions (directions) and speeds in these positions and in the time and distance between consecutive positions. For road map calculations, it is necessary to have information about the path between these positions. If the recorded distance coincides with the length of said guessed curve, it can be considered satisfactory. According to the following values: a starting point and a point of completion of the trajectory, the vector of the speed at the beginning and at the end, the distance, the time needed to travel this road, it is possible to provide a step to guess the path between these points. In accordance with the present invention, it is possible to guess or calculate the trajectory by means of a third-order Bezier curve. The starting point and the termination point are fixed and are the first control point and the last control point, as is known in Bezier curve techniques. Next, the intermediate position between two control points is determined. The second control point is obtained from the first, by adding the velocity vector, and the third control point is obtained from the latter by subtracting the velocity vector. Next, the normalized velocity vectors are multiplied with an appropriate factor (for example, velocity [m / s] * time [s] / 3) for the first approximation of these points. Then the length of the curve can be calculated, and if necessary, it can be adjusted (see next section). Adjusting the length of the Bezier Curve This is useful when giving an approximation of the curve with correct directions. The length is an additional factor that is taken into account, if only two degrees of freedom remain - the length of the starting vector and the length of the termination vector. This procedure changes both vectors uniformly, because usually the speeds do not change very abruptly. If the curve is shorter than the real data, it is preferred to extend the speed vectors (control); and if it is longer, it is preferred to shorten the vectors, and repeat the process. When the actual length and the required length are sufficiently close to each other, then the operational sequence can be stopped. However, said adjustment is provided in an iterative manner so that it is possible to obtain the desired result after a certain number of operations. Calculation of the road sections This is the step in which it is possible to calculate the sections or sections of roads between the junctions of the roads. This step focuses on the geometry of the road network. According to the invention, a starting point is arbitrarily chosen, and the operational sequence continues with the basic operation described above together with the measurements, until the measurements are separated. This is a signal of a crossing. It is also considered to continue the section backwards in order to acquire the concrete section between the crossings. Calculation of road crossings The calculation of road crossings is a separate step, as the geometry and topology of the road network present their greatest complication at junctions. In this step, emphasis is placed on the topology. Measurements (records or parts of curves) are assigned to the corresponding road sections. All the measurements that lead from one section of the road to another section of the same meet. They are like a flow from one pipe to another pipe. The basic operation already described is applied to the collected measurements. It refers to connecting only the two existing sections with the newly calculated "flow" section. The same is done for all combinations of two road sections, which are connected to each other by the measurements. The same procedure can be repeated on the resulting graph or carried out on various graphs obtained from different sources (government institutions, road construction companies, etc.), instead of only based on measurements from our system. Global Calculation The global version is more geared towards geometric accuracy. It requires long trajectories (of at least 500 m) within the measurements. It also allows a partial complementation of the graph. Calculation of a road section To begin with, select the starting point and the completion point of the section (section) of the road. Then all the measurements that go from the starting point to the termination point and that have approximately the same length meet. The basic operation, described above, is applied to the gathered data. It is possible to discard a small portion (100 to 500 m) of the section at the end points, in order to avoid less accurate results.

Adding the road section to the existing graph If a road section has already been calculated, you can attach it to the existing graph. Only the sub-sections, which are not included in the existing graph, can be attached. Calculation of the graph On the other hand, it is necessary to repeat the first two steps, until all the measurements have been used. Start by providing a start with an empty graph, and the final result is the graph of the part of the road network, which had been sufficiently covered with measurements. Experimental results The experimental observations show a high accuracy of the method according to the present invention. Of course, the accuracy depends on the number of measurements made. There is a small percentage of topological errors in the calculated graph. The errors appear mainly if there are parallel roads, closer to each other than twice the GPS error (typically 30 meters) of separation, and at complex junctions. They are due to the lack of accuracy of the GPS system and an excessively long pre-established time interval between the records made. It is anticipated that the percentage of errors may decrease when the methodology begins to use compressed measurements (the dynamic time interval between records with the adapted curve) and include the gyroscope in the on-board unit. Also, the speeds and waiting times are quite accurate. Profiling the network On the other hand, a main operational step according to the present invention, may be the identification and profiling of the crossings. It is possible to merge in the form of a more complex structure of a junction, several vertices in the basic graph of the road network, which are connected to each other and are close to each other. The basic graph of the road network is used together with the raw data to analyze the crossings, in order to define the following properties (and eventually also others) of a crossing: the traffic rules (which roads enter the crossing, which exit from the crossing, and which are connected, if there are traffic lights, which roads have priority, etc.), the traffic pattern (which roads are the most important in a crossing, what is the time to cross the crossing) , type of crossing (in X or star type, roundabout, exits (for example, from a road), etc.), how many lanes are assigned to a specific address, etc. The second line data can be used to differentiate the most important roads from less important roads, in order not to distract the driver when navigating in an area that has an excessive number of less important roads. Again, the data is stored as a graph along with additional auxiliary data structures (matrices, etc.). These data would be basically enough to help a driver navigate. On the other hand, the profiling of the roads is provided, see also Figures 2A to 2C. If there are many vehicles that circulate along the same roads, a quantity of statistical data is available, such as the average speed, the average speed at a given moment of the day, etc. This data is used to outline the connection (road), which consists in assigning the following attributes to each connection: street / road direction (one-way, two-way), distance, average speed or average time to traverse the connection (depending on the time of the week, or similar), the validity of the statistical data (to verify if there is enough data available to inform something substantial about traffic on a particular road / connection), quantity average of traffic (relative to other roads), type of road (remove, street, local road, number of lanes, etc.), the time it has been (most recently) used, and possibly also some other attributes. This is done using the graph of the road network and the raw data. The process has been described in greater detail, which precedes when referring to Figures 2A to 2C of the description. A considered advantage is that (thanks to the curves and the speed adapted in the graph) it is possible to provide the speed at each point of the trajectory (travel) of the vehicle. For this reason it is possible to inform about the exact speeds of the vehicles when they cross a cross section of the road. Other quantities (values) can be adapted analogously to the aforementioned example, in terms of the speed according to the present invention. In general terms, this profiling operation can be carried out using stored traffic data, at any time and in any graph (generated manually or from other sources). If it is observed when the roads have been used, it is possible to find the roads that had not been used by (equipped) vehicles for a long time. It is very likely that such roads are no longer in use and can be erased (usually after some verification) from the road map database. This is a very efficient way to detect auxiliary roads (used to build a road or in other works) or other roads that have stopped working (see the chapter on updating the network). The result is a digital system of road network that is correct from the geometric and topological point of view. It contains statistical data that allows a very accurate navigation of a faster trajectory due to the past experience of all the vehicles integrated in the scheme. However, it is necessary to verify this data manually (with specially equipped verification vehicles) to avoid the possibility of proposing forbidden turns to drivers. Verification The digital road system from the previous section should be covered by verification vehicles, equipped with special equipment, to verify that the database (road map) corresponds to the actual road system. Since the road system is already digitized, it is possible to inform the driver exactly which road to travel in order to achieve the lowest possible road to travel. Optimization can be effected by using some of the well-known principles of road optimization (such as the Chinese Postman Algorithm, known in the Theory of Graphics). Of course, it is possible to make some correction manually, before specially equipped vehicles hit the road for verification. This can further reduce the necessary costs. Another reduction of costs is achieved if these vehicles are sent only to those roads and crossings that had been calculated from insufficient data in quantity or accuracy. This is especially useful when changes are detected in the graph of the road network. This represents a considerable advance with respect to other systems in which there is no road system (or at least not somehow "topologically" ordered) before verification. However, if there is a lack of coherence between the road network system and the digital road network, the driver must capture said data in the special equipment, which then proposes a new road. Changes can be permanent or temporary (with less effect on the digital system of the road network). This process makes verification much quicker and more efficient at a cost level. In fact, the verification adds or eliminates some streets (edges in the graph), and changes the connections, the topology of the same. Roads that may never have been traveled by vehicles, they are not necessarily added manually. In case of need, a pair of vehicles of verification crosses the highway a pair of times with the purpose of introducing it in the system. One aspect taken into account is that these vehicles have to verify the height and width of the tunnels or other obstacles, since these data are difficult to acquire (obtain) otherwise. The result is a digital system of roads that can be used for navigation. These vehicles can be equipped with vibration sensors to determine the quality of the road or with other sensors that may not be directly linked to the road network, but gather other useful information, such as mobile network coverage, or similar . Update of the road network graphic Since the on-board devices are sending data continuously, the described process can be repeated several times in correspondence with an update step. The object is to have the ability to detect new sections (sections) or changes in the road network very quickly and verify these same sections through the process described, very quickly. Given that the raw data recently processed turns out to be very likely the same streets and since some of them have proven to be erroneous through verification processes, it is considered to pay attention to them and to label them accordingly to help prevent the process of update unnecessarily send verification personnel. In general terms, said update operation can be used by using said traffic data at any time and on any graph (generated manually or even from other sources). The raw data, sent by the on-board devices, are used for various purposes. The raw data about the vehicle's trajectory are described with curves. For each section of the trajectory, there are the road sections and the corresponding crossings in the database. If it was not possible to find them in the database, this section of the trajectory is marked and saved in memory for the update of the road network. In the above-mentioned step, the similarity between the curves is a problem taken into account. It is provided to find similar sub-sections of the curves in order to have the ability to identify, if a particular vehicle was on a certain road or part of a road, respectively. The sections that are outside the graph are saved in memory and marked accordingly. When an update calculation is initiated the geometry, topology and profiling can be carried out according to said sections according to the present invention. This is an improvement over the methodologies of the state of the art that only detect the changes without any further processing step. Therefore, said approach of the similarity of the curves can be used for other purposes, such as for electronic toll systems, for example in that it allows the exact determination of where the vehicle is exactly located or the recognition of its form in general. The main server compares the received data with the stored traffic information related to the sections of the road. If this information differs substantially, this is cause for warning. Typically this would suggest a traffic jam. If several vehicles send similar information about abnormal traffic on a specific section of the road, the alert is even more convincing. This operation takes place within a couple of minutes. The data is also used for post processing. The first step is to update the statistical information about traffic related to the sections and crossings of the road. The road sections, corresponding to the new data, are in the database, and their information is updated. The road sections also have information about travel times. A regular check (for example once a month) finds roads that are no longer used and that can be omitted from the database (after some verification). On the other hand, the sections or sections of trajectories that did not have corresponding roads in the database, are used for the calculation of new stretches of road, which are then added to the database. Similarity of Bezier curves To compare two curves, for example, it is first necessary to align them. This alignment can include translations, rotations and scaling. The similarity between the curves is calculated from distances (Euclidean or other) between corresponding pairs of control points, according to the present invention. This calculation can be the sum, average, maximum calculation, minimum calculation, etc.; This depends on the nature of the problem. It is true that similar compositions of control points result in similar curves. The opposite is not always true - it is possible to build similar curves with very different control point compositions. The problem is the parameterization. This can be illustrated, for example, by a right curve, a line. The two control points of the line can be placed anywhere on the line, and the curve will have the same shape; only the parameterization will differ. If only the shape of the curve, not its parameterization, is considered, it is possible to re-parameterize the curves before calculating the similarity. The re-parameterization can be done by collecting points on the curves, and then they are matched with another curve (See the chapter on matching a Bezier curve with an ordered set of points). This curve should have basically the same shape as the original. Find similar sub-sections in a couple of curves This procedure is considered for calculating road maps and for updating traffic statistics. A long curve can describe the trajectory of a vehicle. Roads are also described as curves. It is considered to determine when the vehicle was on one or another road, which part of its trajectory corresponds to which road. First, a definition of the similarities of the curves is needed. To compare the curves, you can use: • absolute criterion. In this case, the minimum length of the sub-sections of the curves being compared should be set. The short sections of curves can always be similar, since their control points are close to each other. It is also possible to configure some criteria about the angle between the curves (between the vectors, which connect the starting point and the end point, for example). • relative criterion. The curves can be aligned at the beginning. First, a sub-curve of the second curve is selected, with the end points closest to the end points of the first curve. The following procedure is recursively repeated according to the present invention: If the curves are similar, they are recorded as similar parts. Otherwise, the first curve (in its middle) is divided, and the second curve closest to the division point of the first curve is divided. Both pairs of sub-curves are compared. At the end, the pairs of similar sub-curves are reported. The system described according to one embodiment of the present invention is a very effective way to generate and profile a digital model of a road network. This type of data is taken into account in an era of mass transit. Instead of building a special infrastructure to deal with traffic analysis, the proposed system uses relatively inexpensive equipment for vehicles, which serves other useful purposes (navigation, messaging service, general fleet control), a public network wireless data (GS; UMTS, CDMA) and a special computer system to analyze high volumes of data. This type of principles is extremely useful for developing countries that have rapidly evolving road systems and that otherwise lack sufficient organizational skills to perform complex operations in order to create a digital model or road network. There are many possibilities of how this system could be used as well. The on-board devices are capable of allowing the driver's navigation if they have a user interface, typically a keyboard and a screen. A navigation requirement can be sent to the server, which also has current information, the server sends the results back to the OBU, which presents the results and guides the driver. The most valuable data consisting of continuously updated data of the digital model of the road network, along with traffic statistics, which helps shipping companies to update their products of entanglement, much more quickly. This is true for countries that already have road maps (EU, USA), and especially for countries that have poor digital models of their road networks (Russia, China, India). The profiled road model helps planners of road structures to increase vehicular traffic where it would have the greatest effect. The module includes traffic flow data, not only general, but also for a particular time of a day, for a particular day in the week, etc. A common question would probably be the following: how much time is needed to get from point A to point B? Each trip or trajectory can be described as an ordered set of measurements, curve.

They can be marked with a path identifier. In this case, all measurements (curves) close to point A and all those close to point B are combined. If a measurement (curve) In the first set has the same identifier as a measurement (curve) in the second set, then the path between these two measurements (curves) is extracted. All these sub-sections of extracted trajectories represent the traffic flow from point A to point B. They can be further analyzed. Since the data of encareteramiento is based on statistical data (that are updated on a daily basis), it is a perfect platform for applications of optimization such as optimization of multiple loads and multiple deliveries, delivery in exact time, optimization of the variation of arrival , optimization of the public transport network. In the case where the graph of the road network includes timing details that define the time needed to travel the connections (road sections) of the graph, it is considered to calculate the fastest road on a basis of time details. These details of the weather can characterize the transit depending on the day of the week or in general terms of the time of day, for example. For example, if a user captures the time of departure, the methodology according to the present invention will determine the fastest road and inform the user of the time resulting from the trip, or the like. It is also considered that the user can capture the desired moment for arrival, in such a way that the algorithm will determine and provide the departure moment, etc. This could be achieved in the following way: each connection of the graph should have information attached about how much time is needed to travel according to the time details. When it is desired to establish the fastest road, the elements visited must also include the timing details. A system of this type could easily be modified to function as an electronic toll system. The main advantage is that, if all the knowledge is used, a complete map of the road network would not be needed. The trajectory is measured as curves, and the probability of identifying the correct road through the use of a curve is much higher than the case of using a simple GPS coordinate measurement. If the shape of the curves is compared, the determination of the actual location of the vehicle is much simpler and more accurate. Figure 3 shows the principle of a system according to an embodiment of the present invention. The plurality of vehicles is illustrated by way of representation by two automobiles, which are equipped with suitable on-board devices. Such devices are adapted to, for example, receive GPS signals to determine the geographical location of each vehicle, respectively. According to this method, but not limited thereto, a GPS satellite 300 can be used. Said satellite 300 provides a position signal to each on-board device of said measuring vehicles. The on-board device can store all of the position data, or alternatively, it can periodically send the data to a central server 301 of a certain location 302. The server 301 is suitably equipped with an antenna 303 and of course with means to receive signals from measuring vehicles. All received information can be stored in the memory in the server unit or for example in another suitable storage medium. The methodology according to the present invention can be executed on said server 301 which according to this mode also serves as a work station (computing station). Additionally, it is also possible to increase a database server to support said server 301 in the storage of a large amount of received position data. In this case, the trajectories of both vehicles, receives the denomination of Highway A and of Highway B, and said roads show two crossings (Crossing A). By means of said received information, the server can store all the trajectories from each vehicle respectively. Further, in accordance with the present invention all trajectories from one or more vehicles traveling (being driven) on a similar road, can be averaged to obtain accurate road models. The area 380 shows by way of example a part of a road to which some dimensions such as the length L and the width have been assigned. According to the present invention, all road sections that are part of the road network graphic, can be characterized by parameters such as: width, length, direction, height above sea level, etc. It is possible to insert other additional parameters such as: average speed, category of the road, or similar. The average speed can be defined according to the time of day or according to the day, for example. Additionally, said parameters may include statistical information about the traffic. These statistics may be provided by third parties, for example, which may include information about the bottleneck or even traffic statistics, such as the number of cars or an estimated value, etc. Figure 4 shows an embodiment of an on-board device that can be installed in a measuring vehicle. An edge device of this type comprises a CPU 400 that is adapted to control all of the operations of said device. The CPU 400 can interconnect all the modules or components, respectively, within said on-board device, according to Figure 4. Said on-board device comprises: a removable storage, 425, a receiver of position signals, 405 , further a module for the approximate calculation of the position, 410, a communication interface, 420, and an internal memory module, 415.

Said communication module 420 can be adapted to communicate with the central server by means of a certain data channel. The use of different techniques such as GSM, CDMA, UMTS, TETRA, General radio interface or similar is considered. Figure 6 shows the principle for averaging several trajectories, represented by Bezier curves, for an averaged curve. Each of trajectories A, B and C is described by a Bezier curve approach based on the positional information of data 60. In the illustration according to Figure 6, only the principle of calculation of agreement is illustrated. with invention In fact, the trajectories of each vehicle are almost identical with the actual shape of the road or street under observation, but for reasons of clarity there is a considerable difference between the trajectories. In general, each trajectory of each vehicle can be described by consecutive Bezier curves. These curves usually have different lengths. To obtain the geometry of the axis of the road, it is necessary to provide an average step on said trajectories corresponding to said plurality of measuring vehicles. This means that the shape of the curves depends on the positional data received from said plurality of vehicles. In this embodiment only three trajectories are described, but it is possible to carry out the methodology according to the present invention on a plurality of vehicles. The positional data 60 may include geographical position data (coordinates) of said measuring vehicles, where said coordinates are used to describe the Bezier curves. The mathematical calculations of said Bezier curves were previously described in detail in the sub-section "Bezier curves". In this mode, the positional data is provided on a time basis, and this means that each of the positional data? T will be transmitted in some way from said plurality of vehicles. The timing may vary, and according to the present invention, it is not fixed. Therefore it is considered to choose a great value for this time if the road has no curves, and in those areas where the road has many curves or crossings, it is possible to adapt the time accordingly. That is, the value is decreased, resulting in fine measurements of the shape of the trajectory. After that, trajectories A, B and C can be used to calculate an averaged curve 65 that corresponds to the existing physical shape of the road. According to the invention, it is considered to average a large number of trajectories (Bezier curves) to obtain the desired result. The algorithm according to the present invention allows an effective average of the Bezier curves, and from the point of view of the computing power it is advantageous and economical. Therefore the present invention obtains the automatic calculation of the road network graph, in which the input is usually formed by measurements from many vehicles (included in the present system), but the same methods can be carried out on some other measurements or also on existing graphs of the road network. On the other hand, the invention allows to obtain an automatic profiling of the network, the input being a graph and the raw data. The graph is obtained as outlined in the foregoing descriptive memory (calculated as above, purchased from someone, etc.), and the raw data is usually measurements from the vehicles in the system of the present invention, but could they are somewhere else (for example, road names, speed limits established by government agencies). In another aspect, the procedure basically consists in the copying (and engraving) of the raw data (some of its parameters) on the graph. The shape of the curve is an aspect taken into account for the identification of the corresponding sections of road and trajectory. On the other hand, the automatic update basically corresponds to what was described above, in which the recognition of the sections of the trajectories that do not correspond to any section of road (and vice versa - sections of road that in recent times have not been covered by any vehicle) is of particular importance. When a sufficient amount of them has been collected, it is possible to calculate new parts of the road network graph. One of the aspects is that it is possible to do this on any graph, which means that it is possible to carry out the update (and also the profiling) on existing road graphics, for example for the EU, USA, Japan, etc. Finally, a verification method is provided in which a final approval of the data is covered. The advantage is that the present invention has an approximation (a calculated graph) and that it can optimize the roads for the verification vehicles, which means a substantial reduction in costs.

In addition, a method is provided to find a faster road within the road network graphic. This finding is based on the timing details that are part of the elements of said road network graph. However, a user of a suitably equipped vehicle may use the information provided by the network graphic according to the present invention to determine (find) the temporarily faster road. For example, if a user wishes to arrive at a particular address at a certain time, the methodology according to the present invention will determine and calculate the fastest highway. Said determination is based on the information included within said graph of the road network, which had also been profiled by using timing details. On the other hand, a method is provided to inspect the flow of traffic, recorded by means of said information data, which may be applicable, for example, for the planning of the road infrastructure. That is, the road network graphic, continuously adapted, provides information about the condition of the traffic, and can be used to determine sub-sections and / or crossings of crowded roads, and the like. Although the invention has been described above with reference to embodiments according to the attached drawings, it is evident that the invention is not limited thereto, but that it can be modified in various ways without departing from the scope of the appended claims. It is noted that in relation to this date, the best method known to the applicant to carry out the aforementioned invention, is that which is clear from the present description of the invention.

Claims (5)

1. CLAIMS Having described the invention as above, the content of the following claims is claimed as property: 1. - Method for modeling the graph of the road network, characterized in that it comprises the following steps: receiving information data from a plurality of vehicles , the information data comprising at least positional data of the vehicles; and modeling the graph of the road network according to the data received.
2. 2. Method according to claim 1, characterized in that the information data comprise at least one of the following: vehicle type, vehicle speed, acceleration and similar, of the plurality of vehicles.
3. 3. Method according to claim 1, characterized in that it also comprises the following step: - Automatically calculate the geometry, topology and traffic statistics of the road network.
4. 4. Method according to any of the preceding claims, characterized in that it is comprised of an automatic fusion of the graphs of the highway network) .
5. 5. Method according to any of the preceding claims, characterized in that an automatic profiling of the road network graph is carried out by using the information data. 6. - Method of compliance with any of the preceding claims, characterized in that a verification is carried out by inspection of the graph, preferably by vehicles that inspect the graph of the road network in itself. 1. - Method of compliance with any of the preceding claims, characterized in that the graph of the road network is used for an optimization step during the verification of the graph of the road network. 8. Method according to any of the preceding claims, characterized in that the modeling is based on mathematical techniques for the processing of curves, arcs, polynomials or the like, carried out on the data. 9. Method according to any of the preceding claims, characterized in that the modeling is based on Bezier curve techniques. 10. Method of compliance with any of the preceding claims, characterized in that the mathematical techniques are based on the average of curves. 11. Method according to any of the preceding claims, characterized in that it further comprises extending the lengths of the curves to a desired predefined length. 12. Method according to any of the preceding claims, characterized in that it further comprises adapting the curve to an ordered set of points, the curve corresponding to the trajectory of a particular vehicle. 13. Method according to any of the preceding claims, characterized in that it also comprises calculating the similarity of a given pair of curves. 14. Method of compliance with any of the preceding claims, characterized in that it further comprises carrying out a step of similarity detection to find similar sub-sections of a given pair of curves. 15. Method of compliance with any of the preceding claims, characterized in that the transmission of the information related to the graph of the network to at least one vehicle of the plurality of vehicles is provided. 16. - Method according to any of the preceding claims, characterized in that the compression of the information data related to the plurality of vehicles is carried out. 17. Method according to one of claims 12 and 16, characterized in that the compression comprises matching Bezier curves. 18. Method according to any of the preceding claims, characterized in that the received information data represent a path of at least one vehicle from among the plurality of vehicles, where each path could be described by Bezier curves, the method also comprising: averaging trajectories associated with at least one vehicle. 19. Method of compliance with any of the preceding claims, characterized in that it also comprises the following steps: calculate a first approximation of the road network graph on the basis of the information data received; profile roads and junctions within the first approach, the result being a profiled road network chart; and - carry out a verification of the profiled network. 20. Method according to any of the preceding claims, characterized in that the calculation is based on techniques Bezier curves and / or mathematical techniques to process curves, arcs, polynomials or the like, carried out on the data. 21. Method according to any of the preceding claims, characterized in that it also comprises the following steps: - detect changes in an existing graph of the road network on the basis of the information data received; store the changes; and implement the changes in the existing graph of the road network. 22. Method according to claim 21, characterized in that the implementation is based on statistical information. 23. Method according to any of the preceding claims, characterized in that it also comprises transmitting information related to the network graph, to at least one vehicle of the plurality of vehicles. 24. Method according to any of the preceding claims, characterized in that it further comprises: carrying out the compression step of the information data selectively within the modeling entity and / or within the plurality of vehicles. 25. Method according to any of the preceding claims, characterized in that it also comprises the next step. store the information data. 26. Method according to any of the preceding claims, characterized in that the calculation is based on digital computing techniques to compute values of fixed points. 27. Method according to any of the preceding claims, characterized in that the information data comprise measurement data, and furthermore comprises normalizing the measurement data according to predetermined threshold values. 28. Method according to any of the preceding claims, characterized in that storage is provided after the execution of a compression algorithm, a coding algorithm, an encoding algorithm, or the like. 29. Method according to any of the preceding claims, characterized in that it also comprises detecting the existence of a phenomenon / effect of multiple trajectories, and in this case assigning less weight to the information received during the step of the calculation. 30. Method according to any of the preceding claims, characterized in that it also comprises measuring the dimensions and / or proportions of the roads by means of a position information providing entity located within the plurality of vehicles. 31. Method according to claim 30, characterized in that the entity is a GPS transceiver within the vehicle. 32. Method according to claim 31, characterized in that the GPS transceiver is coupled to a gyroscope and the like. 33. Method according to any of the preceding claims, characterized in that the similarity of the curves is used to determine the position of the vehicle in the road network graph. 34.- A computer program product, comprising sections of program codes for carrying out the operations according to any of the preceding claims, characterized in that the program is executed in a processor-based device, in a terminal device, in a device of the network, in a portable terminal, in an electronic device of the consumer, or in a terminal enabled for mobile communication. 35.- A computer program product, comprising sections of program codes stored in a machine-readable medium for carrying out operations in accordance with any of the preceding claims, characterized in that the program product is executed in a processor-based device, in a terminal device, in a device of the network, in a portable terminal, in an electronic device of the consumer, or in a terminal enabled for mobile communication. 36.- A software tool, comprising program parts to carry out the operations according to any of the preceding claims, characterized in that the program is implemented in a computer program for execution in a processor-based device, in a terminal device, in a device of the network, in a portable terminal, in an electronic device of the consumer, or in a terminal enabled for mobile communication. 37.- A computer data signal characterized in that it is incorporated in a carrier wave and represents instructions that when executed by a processor cause the operations to be carried out in accordance with any of the preceding method claims. 38.- Server device for modeling the graph of the road network, characterized in that it comprises: - a component for receiving information data from a plurality of vehicles, the information data comprising at least positional data of the vehicles; and a component to model the graph of the road network according to the data received. 39.- Server according to claim 38, characterized in that it further comprises: a component for calculating a first approximation of the road network graph; - a component to profile roads and junctions within the first approach, resulting in a profiled road network chart; and a component to carry out a verification of the profiled network. 40.- Server in accordance with the claim 38, characterized in that it further comprises: a component for detecting changes of the graph of the road network based on the information received; - a component to store the changes; and a component to include the changes in the graph of the road network. 41.- Server in accordance with the claim 38, characterized in that it further comprises: - a component for analyzing the graph of the road network on the basis of the information received; and a component to report analysis results to a third party. 42. - Server according to claim 38, characterized in that it further comprises: a component for performing a compression step of the information selectively within the modeling entity. 43.- Server according to any of the preceding claims 38 to 42, characterized in that it further comprises: a component for storing at least the raw information data received from the plurality of vehicles, attributes associated with the raw data, graphic of the road network, and similar. 44.- Server according to any of claims 38 to 43, characterized in that it further comprises: a component for detecting the existence of a phenomenon / effect of multiple trajectories; and also a component, to assign less weight to the information received. 45. Server according to any of the preceding claims 38 to 44, characterized in that it also comprises a component for measuring dimensions of the road by means of a provider information of position located within the plurality of vehicles. 46.- Server according to any of the preceding claims 38 to 45, characterized in that the received information represents a path of at least one vehicle from the plurality of vehicles, where each path is described by mathematical techniques, including the server in addition: a component for averaging trajectories associated with at least one vehicle. 47. - Server according to claim 46, characterized in that the mathematical techniques correspond to at least one of the following: Bezier curves, arcs, polynomials or the like. 48.- System for modeling a graph of the road network, characterized in that it comprises a plurality of server devices according to claim 38 and a plurality of vehicles providing information data. SUMMARY OF THE INVENTION A method, device and system for modeling a road network graph is provided, comprising the steps for receiving information data from a plurality of vehicles, the information data comprising at least position data, and model the road network graph according to the data received.