DE102017205939A1 - Automatic generation of infrastructure data for a route network - Google Patents

Abstract

A method for generating infrastructure data (18), in particular a route network, for at least one vehicle is disclosed, wherein the method comprises the following steps: recording (10) measurement data concerning at least one infrastructure section during at least one journey over the infrastructure to be measured, transmission the recorded measurement data to at least one, the further process steps performing, data processing device, reworking (12) the previously recorded measurement data by means of correction data (13), updating (14) the infrastructure data from the postprocessed measurement data, enriching (16) the infrastructure data with additional Information, wherein the generated infrastructure data (18) enable the autonomous driving operation of the at least one vehicle.

Description

The present invention relates to a method for the automatic generation of infrastructure data, for describing a route network, a computer program product for carrying out the method, an apparatus and a system for carrying out the method.

For the operation of rail vehicles on railway lines and for the management, maintenance and repair of trackside infrastructure, a data-based overview of the entire infrastructure and distance-related information is necessary.

This data is required to be consistent, complete, consistent, with the appropriate level of detail and data quality. Today, rail networks have a variety of infrastructure and distance-related datasets in electronic form that have grown historically, come from different sources with different accuracy, or are flawed and incomplete.

These available data are insufficient for today's and tomorrow's rail vehicle operation and management, maintenance and repair requirements.

Basically, there is the possibility of carrying out test drives for recording the route characteristics. A disadvantage of this method of data collection are the relatively high costs incurred during the test drives. Due to the high number of parameters to be recorded, the vehicles must be equipped with appropriate sensors.

In addition to increasing susceptibility and complexity, this also entails the disadvantage that qualified personnel must accompany the test drive to support the measuring system.

Thus, a procedure is needed, which reduces the cost of the test drives and offers the opportunity to receive a variety of information on a rail network and process.

In addition, such considerations may also be made for other vehicles and route networks based on technology other than rail vehicle technology. Similar issues include, for example, the driving on roads, and the detection of the local infrastructure and their updating.

In the following, therefore, the invention will be described relatively generally in order to ensure applicability of the invention also to road traffic.

The present invention is thus based on the object of providing a method for the automatic generation of infrastructure data which describe a route network, a computer program product for carrying out the method, an apparatus and a system for carrying out the method.

This object is achieved by means of the main claim and the independent claims. Advantageous developments are the subject of the dependent claims.

According to the invention, the method for generating infrastructure data, in particular infrastructure data which describe a route network, comprises the following steps:

In a first step, measurement data for at least one route segment is acquired during at least one journey over the infrastructure to be measured. Subsequently, a transmission of the previously recorded measurement data to at least one, the further method steps performing device. In a further step, the post-processing of the previously recorded measurement data is carried out with the aid of correction data. Subsequently, the infrastructure data is generated from the post-processed measurement data. Finally, the infrastructure data generated in this way is enriched with additional information, which can be accessed, for example, by additional recording or by other data sources. For example, information about signals, stops, or objects on or on the route is added to the generated infrastructure data.

Preferably, the recording of the measurement data is carried out during at least one, especially made for this test drive, this trip is undertaken by a specially equipped vehicle.

In another advantageous embodiment of the invention, the measurement data are recorded by normal vehicles, which are usually used on this route and are equipped to take over the above-mentioned recording of the measured data. This has the advantage that no special test drives have to be undertaken, but the measurement data can be recorded during normal operation, which saves costs and capacity on the track.

The recorded measurement data is preferably data that is suitable for this purpose are to describe the infrastructure to be surveyed.

Preferably, the measurement data are in the form of suitable coordinates, which reflect the movement patterns of the vehicle on the corresponding infrastructure.

Preferably, the coordinates describe the movement of the vehicle three-dimensionally, which in addition to the progressions of the corresponding infrastructure also describes information about the geographical height of the corresponding infrastructure.

The result of this recording are preferably trajectories. A trajectory preferably describes, by means of the above-mentioned coordinates, the path traveled by a vehicle during the recording of the measured data. Trajectories are thus suitable for describing the infrastructure used by means of the above-mentioned coordinates. Of course, there is a possibility that the trajectories will not cover all possible routes of a route network, since the test drives have not been performed over every possible infrastructure section of the route network. Therefore, the process must be supplemented by further detection steps. These trajectories are usually associated with measurement errors or inaccuracies, which must be corrected in the further process. Trajectories from different test drives, which describe the same infrastructure section, are therefore not necessarily correlated. This is a problem especially in the identification of single-track or single-track and multiple parallel infrastructure sections.

Preferably, the measurement data recorded in this way are transmitted to at least one data processing device which carries out the further method steps.

Preferably, this device is designed to perform planning, documentation and administrative tasks of the route network.

For this purpose, this device is preferably provided stationarily in a planning office for route management.

Preferably, the device is designed to receive the previously recorded measurement data via a suitable connection. For example, this can be done via an internet- or radio-based data connection with the vehicle, or by means of portable data carriers, such as USB sticks.

The post-processing of the previously enumerated measurement data is preferably carried out by means of correction data, which are made available, for example, by a stationary observation station whose coordinates are known. This allows, for example, an offset correction of the present trajectories.

Preferably, in the presence of a plurality of data, which were determined by a plurality of measuring runs, in the post-processing an averaging or filtering, whereby Trajektorienverläufe individual measurements can be corrected, or even measuring trips are excluded from further processing.

Preferably, the recorded measurement data have information that describes the detection time. Measurement data that were recorded, for example, before a known track modification can be immediately excluded from further processing.

Preferably, a new route graph is then subsequently generated from the post-processed measurement data which describe the route network to be measured. This is done by merging the previously measured sections.

In a final step, an enrichment of the generated route graph is provided with additional information, this additional information describing specifics of the route, such as speed limits, stops, signals or the like, thereby increasing the information content of the generated route graph and making it available as generated infrastructure data is provided. This information is additionally recorded during the measurement, or from other data sources, such as a reference data set opened.

The result of this procedure is a database, which is formed by the generated infrastructure data and forms a digital map of the measured route network.

This digital map of a route network is the basis for the operation of rail vehicles on railway lines and also for the management, maintenance and repair of trackside infrastructure necessary.

The method described can also be used for updating a route network analogously to the method already described. This is necessary, for example, after reconstruction or after the commissioning of new lines.

Below is the description of advantageous developments of the method.

Preferably, the measurement data are in the form of coordinates which have been recorded during at least one journey on the infrastructure section to be measured.

Preferably, the acquisition of the measured data is carried out by detection of satellite navigation data. For example, this data can be captured using GPS devices that are state of the art today.

Preferably, such GPS devices are provided in the vehicles to provide GPS data for location and navigation, or to make the acquired data accessible to other vehicle-mounted devices.

Preferably, in GPS-based measurements, the correction data for post-processing of the measurement data is determined by the tracks, or the coordinates of the GPS satellites are detected by a stationary observation station whose coordinates are known. Thus, the relative coordinate information between the GPS satellites and the recording vehicle can be corrected by comparing these coordinates with those between the stationary observation station and the GPS satellites. This correction also allows an indication of absolute GPS coordinates.

Preferably, the observation station is located in the vicinity of the infrastructure section to be measured, in order to achieve the highest possible accuracy of the correction data.

In a further advantageous embodiment, the measurement data is acquired by means of an inertial navigation system (INS), which records the measurement data by means of inertial forces. The main advantage of an INS is that it can be operated without reference, ie independent of any localization signals from the environment. Thus, a collection of infrastructure data can also be done without an existing GPS connection. Thus, it is also possible to detect infrastructure sections where no GPS connection can be made, such as a GPS connection. Tunnel.

A disadvantage arises in this form of measurement, the systemic drift of the sensors, which must be compensated in a suitable manner.

Preferably, the drift is kept low by suitable sensors, or corrected by suitable correction models in the post-processing of the measured data.

Preferably, trajectories can be formed from the recorded measurement data, which describe the traveled infrastructure section. Several trajectories then form a trajectory crowd. Thus, a graphical representation of the test drive is possible.

Preferably, the measurement data are supplemented by further dynamic information from the at least one journey.

Dynamic information describes how the corresponding infrastructure section was traveled. For example, these contain information about the direction of movement, the course of the speed, accelerations and the like.

Preferably, this dynamic information is detected in each spatial direction, whereby additional statements can be made in the further processing of the measurement data.

By means of this dynamic information, it is preferable to be able to identify striking sections or discontinuities in the course of the corresponding route section.

Preferably, the measurement data also contain information from an optical detection of the infrastructure section. This may be, for example, the detection by means of cameras or radar sensors.

Preferably, in this way the detection of objects and elements specific to the driving operation, such as e.g. Switches, signal systems, stations, signs, tunnels, bridges, parallel tracks, crossing points and the like.

Preferably, during the enrichment with additional information, the detection also takes place, which is designed to determine striking sections of the route.

By matching the information content of several post-processed measurement runs, these runs usually being made in both directions of the track, the recognition is preferably capable of identifying distinctive road sections such as intersections, points areas, parallel sections or landmark objects.

A crossing point represents a point on which sections intersect, However, vehicles can not change from one to the other section.

The identification of a crossing point of two route sections preferably takes place in that the trajectories of at least one journey on a first route section are compared with those of a trip on a second route section.

Preferably, a crossing point can be determined by means of the intersection angles of two trajectories, wherein the intersection angle of a crossing point is as obtuse as possible to distinguish it from a point. Thus, if the cutting angle exceeds a certain value, the observed road section is a crossing point.

In another advantageous embodiment, a crossing point is identified by means of the trajectory courses before and after a supposed crossing point. For example, a crossing point can be distinguished from a switch in that intersecting trajectories before and after a crossing point do not converge. In a switch, the trajectories on one side of the switch are identical and lie on each other.

Preferably parallel track sections can be identified by sufficiently resolved measurement data from the trajectories.

In another advantageous embodiment, the identification of parallel running sections takes place with the inclusion of optically recorded information, for example a video associated with the trajectories, whereby an additional, parallel running section can be recognized by means of an automatic image recognition or by a user. This procedure has the advantage that a route section with several parallel route sections can thus be detected with only one trip in one direction if the further parallel route sections were detected by the video. Other prominent infrastructure sections, such as intersections or switches, which are rarely used, for example, and thus not recorded with certainty in the measured data can be detected in this way.

This procedure can preferably also be used for the detection of other significant sections of the route. Thus, the detection of intersection points, switches or other objects can be done about it which also here the need for required trips can be significantly reduced.

Preferably, the infrastructure data generated in this way are suitable for ensuring the autonomous operation of vehicles, in particular rail vehicles, on the measured infrastructure sections, since they record and make available the information required for this purpose.

In a further advantageous embodiment of the invention, the method is in the form of a computer program product with program code stored on a data carrier for passing on and carrying out the method described above.

In this case, the program code may preferably be stored on a USB stick or another suitable data carrier or on a server, from which the program code can preferably be transferred to a corresponding device with a suitable data connection.

In a further advantageous embodiment, the invention is in the form of a device for carrying out the method described above, the device having interfaces which are adapted to detect the measurement data independently, and / or by means of further devices, the device being designed for this purpose to be installed in a vehicle.

Preferably, the device has interfaces that are designed as suitable sensors to detect the measurement data independently. These may be, for example, sensors for detecting position, speed and acceleration data, video cameras or radar sensors.

In another advantageous embodiment of the invention, the device has interfaces which are designed to connect the device with suitable sensors, which enables the device to acquire the required data.

Preferably, a device is also conceivable, which partly has interfaces which are designed as suitable sensors to detect the measurement data independently and partly has interfaces which are designed to connect the device with suitable sensors, whereby the device in the Being able to capture the required data.

In a further advantageous embodiment of the invention, a system is provided which has at least one device for acquiring the measurement data and this at least one device has the acquired measurement data at least transmitted a further device for further processing of the method.

Preferably, the at least one device for detecting the measurement data is designed to be installed in a vehicle.

The at least one device preferably has interfaces which are designed as suitable sensors in order to detect the measurement data independently. These may be, for example, sensors for detecting position, speed and acceleration data, video cameras or radar sensors.

In another advantageous embodiment of the invention, the at least one device has interfaces which are designed to connect the device with suitable sensors, which enables the device to acquire the required data.

Preferably, a device is also provided, which partly has interfaces which are designed as suitable sensors to detect the measurement data independently and partly has interfaces which are designed to connect the device with suitable sensors, whereby the device in the Being able to capture the required data.

In the following, the description of the invention with reference to preferred embodiments with reference to the accompanying drawings.

• 1 ein Flussdiagramm einer Ausführungsform des erfindungsgemäßen Verfahrens, wobei das Verfahren eine erstmalige Erfassung eines Streckennetzes durchführt.
• 2 ein Flussdiagramm einer Ausführungsform des erfindungsgemäßen Verfahrens, wobei das Verfahren eine Aktualisierung eines bestehenden Streckennetzes durchführt.
• 3 eine exemplarische Ermittlung einzelner markanter Streckenabschnitte aus bestehenden Messdaten.

In detail shows:

• 1 a flowchart of an embodiment of the method according to the invention, wherein the method performs a first detection of a route network.
• 2 a flowchart of an embodiment of the method according to the invention, wherein the method performs an update of an existing route network.
• 3 an exemplary determination of individual marked sections from existing measurement data.

1 shows a flowchart of an embodiment of the method according to the invention, wherein the method performs a first detection of a route network.

In step Record 10 the acquisition of the measurement data concerning an infrastructure section to be surveyed takes place. The detection can take place during a journey over the infrastructure to be surveyed, which is undertaken especially for this purpose, or the detection takes place during the normal operation of a vehicle, which in any case moves on the corresponding infrastructure.

After recording 10 The measurement data is transmitted to postprocessing these measurement data 12 ,

The measured data, which is preferably present as a set of trajectories, that is to say as a set of multiple trajectories, from a plurality of measuring runs, are used in the post-processing 12 including correction data 13 reworked. Possible position errors are corrected by means of a fixed reference system or possible sensor deviations are corrected.

The correction data 13 For example, in the case of a GPS-based measurement, it may be obtained by, for example, a stationary observation station which makes it possible to correct the relative coordinates between the GPS satellite and the vehicle.

In the case of inertial measurement methods, the correction data 13 For example, come from a suitable computer model that corrects the drift of the measurement.

This will then be followed in the Update step 14 creates a trace graph, which goes back to at least one measurement. If several measurements are available, in this step the route graph is created by fusing the individual post-processing test drives. Result of the update 14 is a route graph that describes the existing route network completely with all possible ways.

This route graph is then enriched in a further step 16 subjected. In this case, the route graph is automatically enriched with additional information which has been detected by means of a recognition and / or by means of additional sensors. There is also the possibility that users enrich the route graph with additional information by manual input.

Preferably, the enrichment takes place 16 by means of specific information such as turnouts, signal systems, stations, signs, tunnels, bridges.

For example, the recognition adds information about significant infrastructure sections to the route graph, such as intersection points, points or parallel sections. In addition, the possibility exists of an analysis of additional speed and acceleration characteristics from the measurements of discontinuities or to identify possible implausibilities in the route. For example, it is conceivable to plausibilize the height profiles determined from the acquired position data via the speed and acceleration profiles.

From visually recorded additional information, the route graph is supplemented at this point by further information, such as stops, signals or objects on or at the track.

The final result is an updated digital map of a route network containing the required infrastructure data 18 includes, which are suitable for further operation. This flow chart describes the initial creation of such an updated map. However, the method is also suitable for existing infrastructure data 20 to update by means of further measurement data.

2 FIG. 2 therefore shows a flow chart of an embodiment of the method according to the invention, the method updating existing infrastructure data 20 performs.

At the same time, the process remains out of order 1 unchanged, only one additional step is added. This is how the update is done 14 of the route graph incorporating existing infrastructure data 20 ,

For this purpose, the newly determined course of the infrastructure with the existing infrastructure data 20 adjusted and noted differences, for example in the form of coordinate differences.

The existing infrastructure data 20 can be used as the basis of the infrastructure data to be generated 18 be used, the update can also be limited to only a part of the route network, which has been changed, for example, locally limited due to construction.

The recording step 10 In this situation, it can be carried out in such a way that only this eligible section of the line is measured and subjected to an update in the further course of this method.

In addition, the comparison with existing infrastructure data is suitable 20 to make the newly acquired measurement data plausible. If, for example, it is known that a section of the route has not changed, a plausibility check of the acquired measured data by means of a comparison of the existing infrastructure data can be carried out by means of a measuring journey over this section 20 take place, whereby a review of the installed sensors is possible.

3 shows an exemplary determination of individual significant sections of road from existing measurement data.

In the upper part shows 3 a typical infrastructure situation represented by four trajectories. These are two parallel infrastructure sections, which run from top right to bottom left. In addition, in the course of these infrastructure sections, another section branches off by means of a switch, which in the following runs to the bottom right. Again, this section is formed from two parallel infrastructure sections.

The constellation of trajectories shown here corresponds to that after post-processing 12 or the update 14 , Based on this constellation is now the enrichment 16 of the route graph by means of recognition with additional information.

In the lower part shows 3 how the detection independently identifies distinctive infrastructure sections.

First, the subdivision of the recorded trajectories into several sections, A, B, C, D, E and F.

A vehicle that drives from top right to bottom left will change to section A move and then to section C change, so drive on a straight line through the corresponding infrastructure area. However, this vehicle also has the option of section A on section e switch. This is done by a switch 32 allows, which at the end of section A located.

Preferably, driving on the sections takes place B . D and F in reverse to the sections A . C and e instead of. A vehicle thus either moves from section D or section F out on section B to, and reaches this by crossing a turnout 33 in the area in the section D and F converge.

The detection is preferably designed such that it sets the course 32 . 33 can recognize by means of an analysis of the trajectories. These can preferably be done by matching the converging trajectories. A switch 32 . 33 is therefore characterized solely by the fact that on one side of the switch 32 . 33 only one trajectory is present, as can be seen in the sections A and B , on the other side of which the soft two sections each are present, as in this example the sections C and e respectively. D and F ,

Due to the fact that the trajectory consisting of the sections A and e the trajectory from the sections B and D must be cut, the detection is designed such that it can recognize in the intersection of the trajectories, if it is at the intersection of a crossing point 30 is.

For this purpose, the recognition can, for example, use additional altitude information which is available during recording 10 were recorded with. If the height information is the same for both trajectories, so that both trajectories are at the same height at this point, it is very likely to be a crossing point 30 , which is considered as such in the enrichment of the route graph. In addition, a confirmation of the crossing point 30 by means of optical information, such as camera images, which have captured the corresponding section, take place.

Preferably, the recognition also has a possibility for image analysis, so that such a crossing point 30 automatically detected.

The detection of such crossing points 30 is crucial for the continued use of the generated infrastructure data, for example to ensure autonomous operation on the route. A crossing point 30 represents a risk point in this regard, since vehicles can actually collide with each other here. He must therefore be documented beyond doubt.

On the other hand, if the height data of both trajectories do not agree, then it is most likely an underpass in which one section of track is passed underneath the other.

Preferably, an underpass can also be confirmed or identified by optical information.

The trajectories consisting of the sections A and C respectively. B and D run parallel in this illustration. Preferably, vehicles of section A to C or section D moved to B. The trajectories are essentially parallel, with vehicles being able to move past each other in opposite directions. They thus form parallel sections of track 34 ,

The detection of parallel sections of track 34 In contrast to single-track sections on which vehicles move both in one direction and the other, it is essential for the safe operation and safe use of the generated infrastructure data, as this can be used to detect a risk of collision.

The detection of such parallel sections 34 can be done for example by means of the analysis of the position data. If these are present in a sufficiently accurate resolution, then the course of the trajectories alone can lead to two parallel sections of track 34 getting closed.

If such an analysis, due to inaccurate position data, not be possible, the detection can turn optically detected information to detect parallel running sections 34 use.

The invention is not limited only to the embodiments described above. Rather, further embodiments are conceivable, which result from a combination of at least two of the embodiments described above.

In addition, the order of the individual method steps may preferably change, so that, for example, a creation of the line graph is performed only after the enrichment.

LIST OF REFERENCE NUMBERS

10

Record

12

post processing

13

correction data

14

To update

16

enrich

18

infrastructure data

20

existing infrastructure data

30

intersection

32

switch

33

switch

34

parallel sections of track

A

stretch

B

stretch

C

stretch

D

stretch

e

stretch

F

stretch

Claims (14)

Method for generating infrastructure data (18), in particular describing a route network, comprising the following steps: Recording (10) measurement data relating to at least one infrastructure section during at least one journey over the infrastructure to be surveyed, Transmitting the recorded measurement data to at least one data processing device performing the further method steps, Reworking (12) the previously recorded measurement data with the aid of correction data (13), Updating (14) the infrastructure data from the post-processed measurement data, Enrich (16) the infrastructure data with additional information. Method according to Claim 1 in which during the enrichment (16) a recognition step is carried out, which serves to determine striking sections of the route. Method according to Claim 1 or 2 in which at least one trajectory of the traveled infrastructure section is inserted into the measurement data. Method according to one of Claims 1 to 3 wherein the measurement data from satellite navigation data, and / or inertial navigation data are detected, and / or dynamic information is input, and / or information from an optical detection of the infrastructure section are entered. Method according to one of Claims 1 to 4 wherein the correction data (13) are provided by at least one stationary observation station, wherein precise information on the position of this observation station are known, wherein at least one measurement run is corrected by comparison with these correction data (13). Method according to one of Claims 4 or 5 wherein the detecting step of discriminating adjacent infrastructure sections processes the information from optical detection of the route. Method according to one of the preceding claims, wherein the recognition step comprises distinctive sections, such as Crossing points, soft areas parallel sections of track Identified objects. Method according to one of the preceding claims, wherein the measurement data are provided with a detection time, and / or measurement data are processed with different detection times, and / or measurement data are excluded due to their detection time of the further processing. Method according to one of the preceding claims, wherein the infrastructure data (18) generated in this way are used to update existing infrastructure data (20) of the route network. Method according to one of the preceding claims, wherein the at least one vehicle is a rail vehicle. Method according to one of the preceding claims, wherein the generated infrastructure data (18) enable the autonomous driving operation of the at least one vehicle. Computer program product with program code stored on a machine-readable carrier for performing the method according to one of Claims 1 to 11 , Apparatus for carrying out the method according to one of Claims 1 to 11 wherein the device has interfaces that are adapted to detect the measurement data independently, and / or by means of further devices, wherein the device is designed to be installed in a vehicle. System for carrying out the method according to one of Claims 1 to 11 comprising at least one device for acquiring the measured data, wherein the at least one device has interfaces which are adapted to detect the measured data independently, and / or by means of further devices, wherein the at least one device is designed to be installed in a vehicle be, wherein the measurement data for further processing of the method according to one of Claims 1 to 11 be transmitted via a suitable connection to at least one other device.

Images