COMPUTER-IMPLEMENTED METHOD FOR DETERMINING THE POSITION OF A FOLLOWING VEHICLE

Description

[0001] The invention relates to a computer-implemented method for evaluating the accuracy of a swarm trajectory position of a swarm trajectory on a defined road section, defined by a processing device, a computer-implemented method for controlling a following vehicle on a defined road section, in which the method for evaluating the accuracy is used, a control system for controlling a following vehicle, and a computer program product designed to carry out the aforementioned methods.

[0002] Various GNSS (Global Navigation Satellite System) receivers are available on the market that can determine your position. Some of these receivers provide not only the desired position but also information on the position accuracy, although this value is often inaccurate. Another problem is that different manufacturers of these receivers often use different methods to calculate accuracy, making these methods incomparable.

[0003] WO 2020 / 058 463 A1 discloses a method for providing a specified route for a route system of a vehicle, which method comprises the following steps: - providing a plurality of detected trajectories of other vehicles in a route section to be traveled, - determining a specified trajectory from the detected trajectories, - determining a deviation zone from the detected trajectories, wherein the deviation zone is determined based on a deviation of at least individual detected trajectories from the specified trajectory, - determining the specified route at least based on the specified trajectory and the deviation zone.Furthermore, the invention relates to a guideway system for a vehicle, comprising: - a receiving module for receiving detected trajectories in a section of the path to be traveled, - a computing unit to which the receiving module sends the detected trajectories, which is suitable for determining a trajectory specification from the detected trajectories and a deviation zone based on a deviation of at least individual detected trajectories from the trajectory specification, and for determining a guideway specification at least based on the trajectory specification and the deviation zone.

[0004] If, for example, an algorithm for determining a vehicle's position on a defined section of road uses inaccurate values for the accuracy of this position in addition to the received position, for example, to weight data, this leads to inaccurate or even incorrect results. The object of the invention is therefore to propose a method with which values for the accuracy of a position can be provided more reliably.

[0005] This problem is solved by a computer-implemented method with the feature combination of claim 1.

[0006] A computer-implemented method for controlling a following vehicle on a defined road section, a control system for controlling a following vehicle to travel on a road section and a computer program that can carry out the methods are the subject of the independent claims.

[0007] Advantageous embodiments of the invention are the subject of the dependent claims.

[0008] A computer-implemented method for determining the position of a following vehicle on a defined road section comprises the following steps: a) detecting a plurality of ego trajectory positions of ego vehicles moving on the defined roadway section, b) generating a swarm trajectory with a plurality of swarm trajectory positions, wherein the associated swarm trajectory values xi are formed from the plurality of ego trajectory values xn for predefined swarm trajectory values yi, def; c) forming a standard deviation σ i for each formed swarm trajectory value xi of the swarm trajectory;d) storing pairs of the generated swarm trajectory positions and an associated accuracy coefficient for each swarm trajectory position, wherein the accuracy coefficients are proportional to the standard deviations σ i formed for each swarm trajectory position, e) receiving from at least two different sources pairs of a potential position of the following vehicle on the defined roadway section and a source-specific accuracy coefficient associated with the potential position, wherein a first potential position is a swarm trajectory position, and wherein a first accuracy coefficient is proportional to the standard deviation σ i formed for the swarm trajectory position, wherein the swarm trajectory position and the first accuracy coefficient have been generated according to steps a) - d);f) weighting each of the received potential positions using the corresponding source-specific accuracy coefficients; g) determining the position of the following vehicle by fusing the weighted potential positions.

[0009] Swarm trajectories are essentially motion trajectories formed from the fusion of a large number of individual trajectories, with each individual trajectory being assigned to a single vehicle moving on the defined roadway section. These individual trajectories are therefore also referred to as ego trajectories and are associated with the individual vehicles, which are also referred to as ego vehicles.

[0010] According to the method, the swarm trajectory is created from the ego trajectories of the ego vehicles moving on the defined road section, and thus from their GNSS data. The swarm trajectory is essentially formed from a large number of swarm trajectory positions or points (xi , yi ). For each of these points (xi , yi ), the intersection points of the individual trajectories that contributed to the creation of the swarm trajectory at this respective position (xi , yi ) are calculated perpendicular to the direction of travel of the respective ego vehicle. This means that predefined swarm trajectory values yi,def are used to calculate the standard deviation σ i at these points in the direction of travel from the swarm trajectory values xi that are associated with the predefined swarm trajectory value yi,def.The standard deviation σ i is essentially a measure of the dispersion of these values xm around the value xi of the considered swarm trajectory. This standard deviation σ i can be considered a measure of the accuracy that can typically be achieved by GNSS receivers at this considered swarm trajectory position (xi , yi ).

[0011] Once a measure of accuracy in the form of the standard deviation σ i has been determined, the respective swarm trajectory position (xi , yi ) under consideration can then be stored together with an accuracy coefficient. The accuracy coefficient can be the standard deviation σ i itself, but it is also possible to save a representative factor for the standard deviation σ i as an accuracy coefficient. If the standard deviation σ i is not saved directly as the accuracy coefficient, but rather a factor representing the standard deviation σ i, this factor is to be regarded as proportional to the determined standard deviation σ i. "Proportional" does not only mean the mathematical relationship via a constant factor; proportional, in the sense of the method described above, can also mean that the standard deviation values σ i are summarized in groups in order to directly evaluate the accuracy of a position, e.g.Group "high accuracy", "medium accuracy", "poor accuracy".

[0012] The pairs of the generated swarm trajectory position and the corresponding accuracy coefficient are stored together. "Storage" also includes entering them into a map that is made available to a follower vehicle. The follower vehicle is a vehicle that follows all ego vehicles whose ego trajectories were used to form the swarm trajectory.

[0013] The following vehicle can therefore have access to a map created in this way, but the map can also be accessed by other services, for example, which are used for the consolidation of traffic signs present on the defined road section.

[0014] The described method therefore offers the possibility of determining positions or even entire areas with good or poor GNSS accuracy. These accuracies can then be used in other algorithms to estimate the accuracy or a weighting.

[0015] To determine the most realistic position of a following vehicle on a defined road section, data is drawn from two different sources. The first source is the memory, which stores the swarm trajectory position described above along with the associated accuracy coefficient. The second source can be a sensor, which also determines a potential position and outputs an associated accuracy coefficient. Using these source-specific accuracy coefficients for these potential positions, it is now possible to weight the received potential positions and determine the position of the following vehicle from these weighted potential positions.

[0016] The technical advantage of the following vehicle is that the algorithms that determine the position of the following vehicle from the values of one or more sensors now have another source for estimating the accuracy of GNSS data. Knowing the accuracy of the respective sensors is important because weighting takes place when the sensor data is fused. Sensors with higher accuracy are given greater weight. If the following vehicle now has information that the swarm trajectory position can be assessed with a high degree of accuracy, this information can be given a higher weighting than, for example, the potential positions provided by the other sensors. Conversely, however, it is also possible for the other sensors to be given a higher weighting if the GNSS position has a lower accuracy. Overall, this enables improved positioning of the following vehicle.

[0017] The described method therefore makes it possible to correct the problems of erroneous accuracy assessment of commercially available GNSS receivers.

[0018] The following vehicle can therefore also process information from multiple sensors. It is possible to determine a pair of a second potential position and a source-specific accuracy coefficient associated with the second potential position using a sensor assigned to the following vehicle. This means that such a sensor is located in the following vehicle itself, for example, a camera.

[0019] In an advantageous embodiment of the method described above, sensors of the individual ego vehicles moving on the defined roadway section detect the plurality of ego trajectory positions and transmit the plurality of ego trajectory positions to a processing device arranged outside the ego vehicles, whereupon the processing device then generates the swarm trajectory.

[0020] In this advantageous embodiment, the raw data are essentially transmitted to the processing device, so that the processing device generates the swarm trajectory with a plurality of swarm trajectory positions by performing several calculation steps.

[0021] In an alternative embodiment, however, it is also possible for the sensors of the individual ego vehicles moving on the defined roadway section to detect the plurality of ego trajectory positions, and then for each ego vehicle to generate its ego trajectory from its detected ego trajectory positions. Only then does each ego vehicle transmit its generated ego trajectory to a processing device located outside the ego vehicles, which then generates the swarm trajectory from these ego trajectories. In this advantageous alternative embodiment, parts of the computational process for generating the swarm trajectory are performed in the ego vehicles themselves.

[0022] A following vehicle can be controlled on a defined section of road using the following steps: Creating a map for the defined road section with pairs of swarm trajectory positions and associated accuracy coefficients for each swarm trajectory position, as described in the above method, controlling a follow-up vehicle to travel the road section based on the created map.

[0023] GNSS receivers in such follower vehicles do not estimate their accuracy correctly in all situations, although these situations are usually locally reproducible. The map created as described above now contains information about the accuracy of received swarm trajectory positions and thus information about locations where GNSS receivers often estimate their accuracy as too good. If this information from the map is now available to the follower vehicle, the following vehicle can be guided more accurately than previously possible based on this created map.

[0024] The following vehicle can be controlled using a control system of an at least partially autonomous vehicle system. Especially in semi-autonomous or even autonomous driving, it is important to know the reliability of the position data processed to control the following vehicle in order to enable highly precise control of a driverless following vehicle.

[0025] Alternatively, it is also possible for a following vehicle to be controlled by a driver, but with an output unit of a driver assistance system that issues control instructions for controlling the following vehicle based on the created map. An implementation of such a system could, for example, be a navigation system.

[0026] Alternatively or additionally, it is also possible for the second potential position and the associated source-specific accuracy coefficient to be determined using a sensor of an infrastructure in the area of the defined road section. This means that there may also be sensors outside the following vehicle that are arranged on, at, or around the defined road section and are capable of detecting the second potential position of the following vehicle.

[0027] In a computer-implemented method for controlling a following vehicle on a defined road section, as described above, a position of the following vehicle on the defined road section is first determined, and then the following vehicle is controlled to drive on the road section based on this determined position.

[0028] It is possible for the following vehicle to be controlled by a control device of an at least partially autonomous vehicle system. Alternatively, however, it is also possible for an output unit of a driver assistance system to issue control instructions for controlling the following vehicle.

[0029] A control system for controlling a following vehicle to travel along a road section comprises a processing device which is designed to carry out the method for determining a position of a following vehicle on the defined road section as described above, and further comprises a control device for controlling the following vehicle.

[0030] An advantageous computer program product is designed to carry out the method for evaluating the accuracy of the swarm trajectory position of a swarm trajectory on a defined road section defined by a processing device and / or the method for determining a position of a following vehicle on a defined road section.

[0031] Advantageous embodiments of the invention are explained in more detail below with reference to the accompanying drawings. Fig. 1 a schematic top view of a defined road section with several ego vehicles moving along ego trajectories, a swarm trajectory formed from the ego trajectories, and a follower vehicle moving along the swarm trajectory. Fig. 2 a schematic detailed representation of a first advantageous example of the successor vehicle from Fig. 1 ; Fig. 3 a schematic representation of a second advantageous example of the successor vehicle from Fig. 1 ; Fig. 4 a schematic flow diagram in which steps of a method for evaluating the accuracy of a swarm trajectory position defined by a processing device of the swarm trajectory on the defined road section from Fig. 1 are shown; and Fig. 5 a schematic flow diagram showing the steps of a method for determining the position of the following vehicle on the defined road section Fig. 1 represents.

[0032] Fig. 1 shows a schematic top view of a defined roadway section 10 on which several ego vehicles 12 move along their assigned ego trajectories (xn , yn ). Each ego trajectory (xn , yn ) is formed from an infinite number of ego trajectory points, which are composed two-dimensionally of the values xn and yn, where yn are values that represent the direction of travel of the respective ego vehicle 12. The values xn are arranged relative to the values yn on the vertical x-axis (see Cartesian coordinate system in the margin).

[0033] From a multitude of such ego trajectories (xn , yn ), a swarm trajectory (xi , yi ) is formed by fusing the ego trajectories (xn , yn ). This also results in a multitude of swarm trajectory points or swarm trajectory positions (xi , yi ) for the swarm trajectory (xi , yi ). To form the swarm trajectory (xi , yi ), for the sake of simplicity, predefined value positions in the direction of travel yn of the ego trajectories (xn , yn ) that are in Fig. 1 denoted by y 1,def , the x-values of the ego trajectories (xn , yn ), in the example in Fig. 1 xn,1 of the first ego trajectory, xn,2 of the second ego trajectory, and xn,3 of the third ego trajectory are averaged to form an x-value xi of the swarm trajectory (xi , yi ). Accordingly, several x-values from different ego trajectories (xn , yn ) are used, so that it is possible to calculate a standard deviation σ i for the resulting swarm trajectory value xi of the swarm trajectory (xi , yi ) from the majority of x-values.

[0034] In a first in Fig. 1 In the example shown, for the formation of the swarm trajectory (xi , yi ) and the associated standard deviations σ i described above, the ego vehicles 12 send their ego trajectory positions (xn , yn ) to a processing device 18 via corresponding transmitters 16. This processing device 18 receives the ego trajectory positions (xn , yn ) and uses this information to determine the swarm trajectory (xi , yi ) and the respective associated standard deviation σ i in a processing module 20. Depending on the implementation, the standard deviation σ i is treated directly as an accuracy coefficient KG, which indicates the accuracy of the determined swarm trajectory positions (xi , yi ). Alternatively, however, it is also possible to convert the determined standard deviations σ i into a representative accuracy coefficient KG, which is proportional to the standard deviations σ i.Proportional does not just mean a purely mathematical proportionality with a constant conversion factor; it is also possible to group standard deviations σ i into evaluation criteria, which are then treated as an accuracy coefficient KG. Such groups can be, for example, "high accuracy," "medium accuracy," or "poor accuracy."

[0035] The processing device 18 then stores the pairs of generated swarm trajectory positions (xi , yi ) and the respectively associated accuracy coefficients KG in a storage device 22. It is possible for these pairs to be stored in the form of a map in which the accuracy coefficient KG is then plotted for each generated swarm trajectory position (xi , yi ).

[0036] To detect the ego trajectory positions (xn , yn ), the ego vehicles 12, as in Fig. 1 As shown, sensors 24 are provided. These sensors 24 can be cameras, for example; however, it is also possible for the ego vehicles 12 to receive GPS data from a backend, so that the sensors 24 in this case are formed by a corresponding GPS receiver.

[0037] As an alternative to the possibility that all calculation steps are performed in the processing device 18, it is also possible for the ego vehicles 12 to have their own processing modules 20 in addition to their sensors 24, wherein the respective ego trajectory (xn , yn ) is formed in these separate processing modules 20 from the ego trajectory positions (xn , yn ) of the respective ego vehicle 12. The ego trajectories (xn , yn ) thus generated are then transmitted directly to the processing device 18 in order to determine the swarm trajectory (xi , yi ) and the associated standard deviation σ i .

[0038] Once the processing device 18 in the processing module 20 has determined the swarm trajectory (xi , yi ) and the associated standard deviation σ i or the associated accuracy coefficient KG and stored it in the storage device 22, e.g. in the form of a map, it is possible to send this information, e.g. the stored map, to a following vehicle 26 that follows the ego vehicles 12 in time on the defined road section 10. The following vehicle 26 receives the formed swarm trajectory (xi , yi ) and the associated accuracy coefficients KG via a receiver 28 and is then controlled via a control device 30 based on the received map.

[0039] The control device 30 can be part of an at least partially autonomous vehicle system 32, in which the control of the ego vehicle 12 takes place partially autonomously or fully autonomously via a control unit 34, or the control device 30 communicates with an output unit 36 of a driver assistance system 38, which outputs control instructions to a driver of the following vehicle 26, for example via a navigation system display.

[0040] The follower vehicle 26 is shown in a schematic detailed representation of a first advantageous example of the follower vehicle 26 in Fig. 2 shown.

[0041] Fig. 3 shows a schematic detailed representation of a second advantageous example of the successor vehicle 26 from Fig. 1 , in which the control device 30 is designed to determine a position of the following vehicle 26 on the defined roadway section 10. For this purpose, the control device 30 not only uses the swarm trajectory (xi , yi ) received from the processing device 18 and its associated accuracy coefficients KG,i, as already described above, but also uses data from a second source 40 that are related to the position of the following vehicle 26. Accordingly, the control device 30 receives a potential position (x pot , y pot ) of the following vehicle 26 from at least two different sources 40, weights these potential positions (x pot , y pot ) based on their associated accuracy coefficients KG, and then determines the position of the following vehicle 26 by fusion. As in Fig. 3 As shown, the second source 40 can be, for example, a sensor 24 of the following vehicle 26, such as a camera. However, it is also possible for the received information to be processed to originate from a sensor 24 associated with an infrastructure 42 located in the area of the defined roadway section 10. This can also be, for example, a camera that is set up or permanently installed in the area of the roadway section 10.

[0042] Based on the position of the following vehicle 26 thus determined, the control device 30 can then, as in the case of Fig. 2 described first example, control the following vehicle 26.

[0043] Overall, with regard to the Fig. 1 bis 3 a control system 44 is described with which the following vehicle 26 can be controlled more reliably than previously known via the processing device 18 and the control device 30.

[0044] With regard to this control, Fig. 4 a schematic flow diagram is shown which evaluates the steps of a method for evaluating the accuracy of a swarm trajectory position (xi , yi ) defined by the processing device 18. In a first step, a plurality of ego trajectories (xn , yn ) from a plurality of ego vehicles 12 are recorded. In the next step, the swarm trajectory (xi , yi ) is then formed from these ego trajectories (xn , yn ). In the subsequent step, the standard deviation σ i is formed for each formed swarm trajectory value xi of the swarm trajectory (xi , yi ).

[0045] In a further step, pairs are stored, which consist of the generated swarm trajectory position (xi, yi) and an associated accuracy coefficient KG,i. This can be stored, for example, in a map. In a final step, a follow-up vehicle 26 is then controlled based on the map data.

[0046] With reference to a positioning of the following vehicle 26 on the defined road section 10, Fig. 5 a schematic flow diagram with steps of a method for determining the position of the following vehicle 26 on the defined road section 10. In a first step, as with reference to Fig. 4described, a map is created. In a next step, the following vehicle 26 then receives 40 potential positions (x pot , y pot ) with associated accuracy coefficients KG from at least two sources. In a further step, these received potential positions (x pot , y pot ) are weighted on the basis of the accuracy coefficients KG and then, in a further step, merged to form the position of the following vehicle 26. Based on the position of the following vehicle 26 thus determined, the following vehicle 26 can then be controlled via the control device 30. List of reference symbols

[0047] 10Road section 12Ego-vehicle 16Transmitter 18Processing device 20Processing module 22Storage device 24Sensor 26Following vehicle 28Receiver 30Control device 32(Semi-)autonomous vehicle system 34Control unit 36Output unit 38Driver assistance system 40Source 42Infrastructure 44Control system KG Accuracy coefficient σ i Standard deviation (xn , yn )Ego trajectory (xi , yi )Swarm trajectory (x pot , y pot )Potential position yn Direction of travel y 1,def Predefined value position in direction of travel ynxn,1 x-value of the ego trajectory (xn , yn )

Claims

1. Computer-implemented method for determining a position of a trailing vehicle (26) on a defined road section (10), having the following steps: - a) capturing a multiplicity of ego trajectory positions (xn, yn) of ego vehicles (12) moving on the defined road section (10); - b) generating a swarm trajectory (xi, yi) having a multiplicity of swarm trajectory positions (xi, yi), wherein, for predefined swarm trajectory values (yi, def), the associated swarm trajectory values (xi) are formed from the multiplicity of ego trajectory values (xn); - c) forming a standard deviation (σi) for each formed swarm trajectory value (xi) of the swarm trajectory (xi, yi); - d) storing pairs of the generated swarm trajectory positions (xi, yi) and an associated accuracy coefficient (KG) for each swarm trajectory position (xi, yi), wherein the accuracy coefficients (KG) are proportional to the standard deviations (σi) formed for each swarm trajectory position (xi, yi), characterized by the following further steps of: - e) receiving, from at least two different sources, respective pairs of a potential position (xpot, ypot) of the trailing vehicle (26) on the defined road section (10) and a source-specific accuracy coefficient (KG) associated with the potential position (xpot, ypot), wherein a first potential position (xPot1,yPot1) is a swarm trajectory position (xi, yi), and wherein a first accuracy coefficient (KG1) is proportional to the standard deviation (σi) formed for the swarm trajectory position (xi, yi), wherein the swarm trajectory position (xi, yi) and the first accuracy coefficient (KG1) were generated in accordance with steps a) - d); - f) weighting each of the received potential positions (xpot, ypot) based on the associated, source-specific accuracy coefficients (KG); - g) determining the position of the trailing vehicle (26) by fusing the weighted potential positions (xpot, ypot).

2. Computer-implemented method according to Claim 1, wherein, in order to generate the swarm trajectory (xi, yi) having a multiplicity of swarm trajectory positions (xi, yi), sensors (24) of the individual ego vehicles (12) moving on the defined road section (10) capture the multiplicity of ego trajectory positions (xn, yn) and transmit the multiplicity of ego trajectory positions (xn, yn) to a processing device (18) arranged outside the ego vehicles (12), and the processing device (18) generates the swarm trajectory (xi, yi).

3. Computer-implemented method according to Claim 1, wherein, in order to generate the swarm trajectory (xi, yi) having a multiplicity of swarm trajectory positions (xi, yi), sensors (24) of the individual ego vehicles (12) moving on the defined road section (10) capture the multiplicity of ego trajectory positions (xn, yn), each ego vehicle (12) generates its ego trajectory (xn, yn) from its captured ego trajectory positions (xn, yn), each ego vehicle (12) transmits its generated ego trajectory (xn, yn) to a processing device (18) arranged outside the ego vehicles (12), and the processing device (18) generates the swarm trajectory (xi, yi) from the ego trajectories (xn, yn).

4. Computer-implemented method according to Claim 1, wherein a pair of a second potential position (xpot2, ypot2) and a source-specific accuracy coefficient (KG2) associated with the second potential position (xpot2, ypot2) is determined using a sensor (24) assigned to the trailing vehicle (26) or using a sensor (24) of an infrastructure (42) in the region of the defined road section (10).

5. Computer-implemented method for controlling a trailing vehicle (10) on a defined road section (10), having the following steps: - determining a position of the trailing vehicle (26) on the defined road section (10) by performing a method according to Claim 1; - controlling a trailing vehicle (26) so as to drive on the road section (10) based on the determined position.

6. Computer-implemented method according to Claim 5, wherein the trailing vehicle (26) is controlled using a controller (30) of an at least partially autonomous vehicle system (32), or wherein an output unit (36) of a driver assistance system (38) outputs control specifications for controlling the trailing vehicle (26).

7. Control system (44) for controlling a trailing vehicle (26) so as to drive on a road section (10), having a processing device (18) that is designed to perform a method according to either of Claims 1 and 4, and a controller (30) that is designed to control the trailing vehicle (26) according to either of Claims 5 and 6.

8. Computer program product that is designed to perform the methods according to one of Claims 1 to 6.