Method for operating a driver assistance system in a vehicle

Abstract

The invention relates to a method for operating a driver assistance system (2) in a vehicle (1), wherein the driver assistance system (2) is designed to learn a driving trajectory in a learning mode and follow the learned driving trajectory (5) in a following mode (26). In the following mode (26), the method has the steps of: localizing (S1) the vehicle (1) relative to the learned driving trajectory (5); ascertaining (S2) a planned driving trajectory (12, 13), wherein a deviation between the ascertained planned driving trajectory (12, 13) and the learned driving trajectory (5) lies within a specified first tolerance range (24); checking (S3) whether the deviation between the ascertained planned driving trajectory (12, 13) and the learned driving trajectory (5) additionally lies within a second tolerance range (25) which is specified for the localization process and which at least partly deviates from the first tolerance range (24); and, only if this is the case, initiating (S4) the process of following the planned driving trajectory (12, 13) in the following mode (26).

Description

[0001] Method for operating a driver assistance system in a vehicle

[0002] The invention relates to a method for operating a driver assistance system in a vehicle. The driver assistance system is designed to learn a driving trajectory in a learning mode and to follow the learned driving trajectory in a tracking mode. Furthermore, the invention relates to a vehicle, a control device for a vehicle, and a computer program product for implementing such a method.

[0003] A vehicle can have a driver assistance system that is designed to follow a driving trajectory learned in a learning mode in a follow-up mode. In follow-up mode, the vehicle then follows the learned driving trajectory at least with assistance, in particular fully automatically. Such a driver assistance system is, for example, a trained parking assistant that has been taught an entry and / or exit trajectory from a starting location to a destination in a parking environment in the learning mode, so that this trajectory can be followed again at least with assistance in follow-up mode. Alternatively or additionally, the driver assistance system can, for example, be a reversing assistant in which a driving trajectory is followed forward in the learning mode, which is then followed at least with assistance in reverse in the follow-up mode.

[0004] The operation of the driver assistance system is based, among other things, on locating the vehicle in its surroundings, such as in the parking area. For this purpose, environmental information describing the vehicle's surroundings is evaluated and compared, for example, with a map of the surroundings stored in the vehicle, on which the learned driving trajectory is plotted. Based on at least one object in the surroundings that is both described by the environmental information and plotted on the map, the vehicle can then be located relative to the learned driving trajectory and thus localized in the surroundings.

[0005] DE 10 2019 133 967 A1 discloses a method for operating a parking assistance system in a first and a second operating mode. In the first operating mode, a localization device detects an environment, and a relative position of the vehicle is determined based on the detected environment. In the second operating mode, a vehicle dynamics parameter is detected using a vehicle dynamics detection device. A trajectory for a parking maneuver of the vehicle is determined based on the relative position, with the relative position being determined based on an availability criterion for position determination in the first operating mode and based on the vehicle dynamics parameter in the second operating mode.

[0006] It is the object of the invention to provide a solution by means of which a vehicle can be reliably located in a tracking mode of a driver assistance system.

[0007] The problem is solved by the subject matter of the independent patent claims.

[0008] A first aspect of the invention relates to a method for operating a driver assistance system in a vehicle. The driver assistance system is designed to specify, at least in an assisted manner, in particular semi-automatically or fully automatically, a longitudinal and / or lateral guidance of the vehicle. Preferably, it can control the longitudinal and lateral guidance fully automatically.

[0009] The driver assistance system is designed to learn a driving trajectory in a learning mode and to follow the learned driving trajectory in a tracking mode. In learning mode, for example, a driver drives the vehicle manually from a starting location to a destination, during which, for example, environmental information describing the vehicle's surroundings is recorded, evaluated, and / or stored. For example, by evaluating the environmental information, the driving trajectory followed in learning mode can be plotted on a map. In tracking mode, the driving trajectory followed in learning mode is followed at least with assistance, in particular fully automatically. This may require the vehicle to be located or localized on the map and thus relative to the learned driving trajectory.Several learned driving trajectories can be provided, allowing, for example, various parking and / or reversing maneuvers or other types of driving maneuvers to be performed when the driver assistance system is activated. The driver assistance system could be, for example, a trained parking assistant.

[0010] The invention is based at least on the recognition that the case may arise where the driving trajectory to be followed in tracking mode deviates at least partially from the learned driving trajectory, for example, to avoid a temporary obstacle or because the vehicle was manually driven away from the learned driving trajectory by the driver. It may then be provided that a planned driving trajectory is specified instead of the learned driving trajectory. However, this may result in the localization of the vehicle along the planned driving trajectory being of insufficient quality and / or at least partially impossible.This is typically only detected when the planned trajectory is followed, whereupon the driver assistance system, for example, aborts the assisted operation of the vehicle or continues with lower accuracy compared to following the learned trajectory, which should be avoided. It is therefore advisable to operate the driver assistance system in such a way that it can always be assumed that a planned trajectory can actually be followed with continuous and / or reliable localization of the vehicle. It should therefore be checked in advance whether the localization of the vehicle along the planned trajectory will be possible and / or sufficiently accurate.

[0011] The method includes locating the vehicle relative to the learned driving trajectory. Localization is performed by applying a localization algorithm to environmental information describing the vehicle's surroundings. The environmental information can alternatively be referred to as sensor information. It is acquired, for example, by a sensor device of the vehicle and then transmitted to a control device of the vehicle, which preferably carries out the localization of the vehicle and the further method steps. The environmental information describes, for example, an image of the surroundings. The environmental information is preferably acquired continuously during operation of the driver assistance system in the vehicle.In addition to the environmental information, driving information describing a vehicle's journey, such as speed, distance traveled, and / or steering angle, can be recorded and taken into account by the localization algorithm. The driving information can also have been recorded in learning mode and stored for the learned driving trajectory. The localization algorithm includes, for example, at least one regulation and / or rule, the implementation of which results in localization. The localization algorithm can alternatively be referred to as a localization criterion. The localization algorithm accesses, for example, a map of the surroundings on which the learned driving trajectory and at least one object in the vicinity of the learned driving trajectory are plotted.The localization algorithm can be based, for example, on simultaneous localization and mapping (SLAM) methods. It is thus known where the vehicle is located relative to the travel trajectory and / or in which orientation it is positioned relative to the travel trajectory. The vehicle is localized, for example, after the driver of the vehicle has manually activated the driver assistance system. Alternatively or additionally, the vehicle can be localized automatically after the vehicle has been located in an area of ​​a starting location of the learned travel trajectory. This can be done, for example, by determining the vehicle's position, in particular based on data from a global navigation satellite system (GNSS).In the following, it is assumed that the tracking mode has already been activated and that the vehicle is therefore being localized. The result of the localization is preferably that the vehicle is, for example, at the starting location or destination of the learned driving trajectory or within a specified area around the starting location or destination.

[0012] The method comprises determining a planned travel trajectory by applying a planning criterion to the learned travel trajectory and the environmental information. The planning criterion can also be applied to the travel information. A deviation between the determined planned travel trajectory and the learned travel trajectory lies within a first tolerance range specified for the planning criterion. Thus, the actual travel trajectory for the vehicle is planned on site, referred to here as the planned travel trajectory. The planned travel trajectory is the trajectory that is actually to be followed in follow-up mode, whereas the learned travel trajectory is an ideal travel trajectory that is followed when operating in follow-up mode without any obstacles between the starting and destination locations and / or without the occurrence of technical problems. Preferably, the planned travel trajectory corresponds entirely to the learned travel trajectory.However, it may be the case that, for example, due to a current position and / or orientation of the vehicle that deviates from a position and / or orientation at the start and / or destination of the learned driving trajectory, the planned driving trajectory deviates at least partially from the learned driving trajectory. Alternatively or additionally, an obstacle along the learned driving trajectory can lead to a local deviation of the planned driving trajectory from the learned driving trajectory. Alternatively or additionally, the planned driving trajectory can differ from the learned driving trajectory, for example with regard to a steering angle in a curve. This can be the case, for example, if the driving trajectory was learned in a different direction than the one intended to be followed in follow-up mode.Basically, by determining the planned driving trajectory, a check is carried out to determine whether the learned driving trajectory can be followed without any changes, or whether at least partial changes to the learned driving trajectory are necessary due to current circumstances.

[0013] The first tolerance range specifies how much the planned trajectory may differ from the learned trajectory. For example, the first tolerance range can specify a maximum distance between the planned trajectory and the learned trajectory. Alternatively or additional specifications can be defined by the first tolerance range. This is a typical procedure, for example, to still be able to use the driver assistance system if the direction of travel changes between the starting point and destination without having to learn the changed direction of travel. Furthermore, this can be used to take into account temporary obstacles such as a garbage can, a pedestrian, a parked vehicle or other objects that are not present when the vehicle is operating in learn mode and can be avoided without the drive trajectory having to be relearned.

[0014] The deviation between the planned and learned driving trajectory within the meaning of the invention can alternatively be referred to as the difference between the determined planned driving trajectory and the learned driving trajectory. The deviation can relate to various parameters that define the respective driving trajectory.

[0015] The method includes checking whether the deviation between the determined planned driving trajectory and the learned driving trajectory also lies within a second tolerance range specified for localization. The second tolerance range deviates at least partially from the first tolerance range. It is therefore determined whether the previously determined planned driving trajectory meets the requirements required for localization, whereby these requirements are at least partially different from the requirements that must be met by the planned driving trajectory. Therefore, the second tolerance range specified for localization is relevant here, defining which deviations are considered acceptable and / or sufficiently small and which are considered no longer acceptable and / or too large.For example, the first tolerance range may specify that the planned trajectory may be no more than 1 meter away from the learned trajectory. However, due to stricter localization regulations, the second tolerance range may specify that, for example, only distances of 80 centimeters between the planned trajectory and the learned trajectory are permitted, for example because otherwise the vehicle's localization would no longer meet a quality requirement due to its distance from the learned trajectory. In such a case, for example, a determined planned trajectory that meets the specifications of the first tolerance range is now recognized as a trajectory that does not meet the specifications of the second tolerance range.

[0016] The respective tolerance range can be defined by at least one limit value, in particular by several limit values. Alternatively or additionally, it can be defined as a percentage, i.e., specifying that the driving trajectories may only differ from each other by 1 percent or another percentage.

[0017] The method comprises permitting the following of the planned driving trajectory in follow-up mode only if the deviation between the determined planned driving trajectory and the learned driving trajectory lies within the second tolerance range specified for localization. Therefore, a check is carried out in advance, i.e., before the planned driving trajectory is executed, to determine whether the specified requirements for localization, which are quantified here by the second tolerance range, are met. This prevents, for example, the planned driving trajectory from deviating from the learned driving trajectory to such an extent that localization can only be performed with low and possibly inadequate quality, or possibly not at all, which could then lead to the driver assistance system's operation being aborted.The specifications of the planning criterion and thus of the first tolerance range are considered insufficient to decide whether, for example, the vehicle should remain in tracking mode, for example, to at least follow the planned driving trajectory with assistance, or not. The probability of a reduction in the accuracy and thus the performance of the tracking mode, as well as the abort of a driving maneuver along the planned driving trajectory, is at least reduced because it has been clarified whether problems or difficulties with localization will arise that were previously identified due to the deviation outside the second tolerance range. Therefore, the vehicle can now always be reliably localized in the driver assistance system's tracking mode.

[0018] The method can be carried out entirely by the vehicle's control device. The method can be understood as a computer-implemented method. One exemplary embodiment provides that, if the deviation between the determined planned travel trajectory and the learned travel trajectory lies outside the second tolerance range, an alternative planned travel trajectory is determined and checked. Alternatively or additionally, the follow-up mode is at least paused, in particular deactivated. This actively disallows following the planned travel trajectory if the deviation lies outside the second tolerance range. For the alternative planned route, for example, the method described above is at least partially carried out again. A further planned travel trajectory is then determined, which is referred to here as the alternative planned travel trajectory.The alternative planned trajectory deviates at least partially from the previously determined planned trajectory. Only if the alternative planned trajectory was planned in such a way that the deviation between the determined alternative planned trajectory and the learned trajectory lies within the second tolerance range is the following of the alternative planned trajectory permitted and can, for example, be carried out. Ultimately, this demonstrates how reliably the system always responds to deviations that lie outside the second tolerance range.

[0019] It can be provided that before the alternative planned travel trajectory is determined and / or before the follow-up mode is paused, in particular deactivated, a waiting period is observed and / or a change in the position and / or orientation of the vehicle occurs. This makes it possible to react to obstacles that are only temporarily present, such as a pedestrian, if the obstacle moves within the waiting period in such a way that at least small deviations from the learned travel trajectory are sufficient to avoid the obstacle or, if necessary, the obstacle no longer needs to be avoided. The change in position and / or orientation can help to move the vehicle closer to the learned travel trajectory, for example, thus reducing the deviation. This makes it clear how, if necessary, following in follow-up mode can still be permitted using the alternative planned travel trajectory.The movement to the new position and / or orientation is at least assisted, preferably fully automatic, i.e. it can be planned and, in particular, carried out by the driver assistance system.

[0020] A further embodiment provides that the respective tolerance range specifies an area around the learned travel trajectory, within which the planned travel trajectory may run. The area can, for example, be designed as a tube around the learned travel trajectory. Typically, a width of the area perpendicular to a longitudinal direction of the travel trajectory is fixed. The width can, for example, be 0.5 meters, 1 meter, 1.5 meters, 2 meters, 3 meters, 5 meters or in particular 10 meters. Other widths than those mentioned, in particular widths between the widths mentioned, are possible. As soon as the learned travel trajectory leaves the area, a distance or a separation between the learned travel trajectory and the planned travel trajectory is too large to lie within the respective tolerance range. The area can therefore be described by the first tolerance range and / or the second tolerance range.The tolerance range can vary locally along the learned driving trajectory, meaning it can be locally wider than at other points along the driving trajectory, for example. Such a range is particularly useful in the context of determining the planned driving trajectory, as it can determine whether the planned driving trajectory is close enough to the learned driving trajectory. In the context of localization, this range is suitable, for example, to ensure that a perspective on the environment has changed only to the extent that objects required for localization can still be identified.

[0021] A further embodiment provides that the respective tolerance range specifies a permitted maximum change in the orientation of the vehicle according to the planned travel trajectory compared to the orientation of the vehicle according to the learned travel trajectory. Such a limitation of the maximum change in orientation is justified, for example, by the fact that an opening angle of the sensor device is fixed and thus, if the change in orientation is too great, the environment is no longer detected in such a way that there is sufficient overlap with the environment as detected in the learning mode. The permitted maximum change in orientation can be set to 30 degrees, for example. Alternatively, the maximum change in orientation can be 10 degrees, 20 degrees, 40 degrees, 50 degrees, or 60 degrees. Smaller, larger, or changes lying between these values ​​are possible.The maximum permissible change may depend on the sensor device that provides the environmental information on which the localization and determination of the planned driving trajectory is based. This ensures that the determined planned driving trajectory always provides sufficient meaningful information regarding the localization, as well as the environmental information provided at the respective location along the planned driving trajectory, to enable the driver assistance system to operate.

[0022] According to an additional exemplary embodiment, the respective tolerance range specifies a permitted maximum distance for which the planned travel trajectory may deviate from the learned travel trajectory. For example, it can be specified that the planned travel trajectory does not run along the learned travel trajectory only for 2 meters, 3 meters, 5 meters, 10 meters, 15 meters or, in particular, 20 meters. The route begins, for example, at a last point of the planned travel trajectory that lies on the learned travel trajectory and ends at a first point of the planned travel trajectory that again lies on the learned travel trajectory. By specifying the maximum distance, it is ensured that, for example, if localization should be at least partially difficult or not possible along the planned travel trajectory, the last localization performed does not lie further back than the permitted maximum distance.The permitted maximum distance is selected, for example, such that the vehicle can be positioned along the route by evaluating the driving information. Preferably, at least the permitted maximum distance according to the second tolerance range differs from the permitted maximum distance according to the first tolerance range. If the permitted maximum distance differs, the area and / or the maximum change in orientation can be selected to be the same for the first and second tolerance ranges, or vice versa. Finally, three possible parameters or values ​​were mentioned with respect to which the tolerance ranges can differ from one another. The method is therefore particularly versatile.

[0023] A preferred embodiment provides that the second tolerance range is specified at least partially dependent on the applied localization algorithm. Depending on how precisely the localization is performed, for example, whether it is based on camera data as environmental information and / or radar data as environmental information, different second tolerance ranges can be specified, since, for example, different influences on the localization of the vehicle are expected. The second tolerance range should be selected depending on the conditions under which the localization algorithm can reliably and / or precisely localize the vehicle. Ultimately, this makes it possible to quantify when the quality of the localization is sufficiently good to permit following the planned travel trajectory in follow-up mode.

[0024] Furthermore, one exemplary embodiment provides that the respective tolerance range is at least partially dependent on the number and / or arrangement of sensor devices in the vehicle. The vehicle has, for example, a plurality of sensor devices, each of which detects at least part of the environment and thus provides the environmental information as a whole. Depending on how many sensor devices are present and / or where the individual sensor devices are arranged, the area, the maximum change in orientation and / or the permitted maximum distance can be selected to be of different sizes. The more sensor devices are provided, in particular the greater the density of sensor devices in the vehicle, the larger the area, the maximum change in orientation and / or the permitted maximum distance can be selected.

[0025] Alternatively or additionally, the respective tolerance range depends at least partially on the quality of the environmental information. The quality of the environmental information is influenced, for example, by contamination of the sensor device, light incidence, such as at least local glare on a camera acting as a sensor device, and / or depends on the accuracy of the sensor device. The respective tolerance range can therefore depend on the sensor device used to acquire the environmental information, on the basis of which localization takes place and the planned travel trajectory is determined. Thus, factors influencing localization and the determination of the planned travel trajectory are taken into account, and these therefore influence the respective tolerance range.

[0026] According to another exemplary embodiment, the environmental information comprises data from at least one camera, a radar device, and / or a LiDAR device. The camera, the radar device, and / or the LiDAR device are then the sensor device that captures and provides the environmental information. It is assumed here that the environmental information has been transmitted to the vehicle's control device, i.e., that the environmental information is available to the control device. The respective sensor device is preferably arranged in a front area, a rear area, and / or a side area of ​​the vehicle. For example, the sensor device is a front camera, a rear camera, a side camera, and / or a surround-view camera.Alternative or additional data are possible, such as ultrasound data or other environmental data that describe a distance to an object in the environment and / or an image of the environment and / or at least one object in the environment.

[0027] In addition, one embodiment provides that the determined planned travel trajectory only deviates at least partially from the learned travel trajectory if the vehicle is manually moved away from the learned travel trajectory. For example, at the beginning of operation in follow-up mode and / or during operation in follow-up mode, a manual movement away from the learned travel trajectory can occur, for example, due to an obstacle that the vehicle driver manually avoids. Due to such a manual movement away from the learned travel trajectory, a travel trajectory that deviates at least partially from the learned travel trajectory can be determined when determining the planned travel trajectory.

[0028] Alternatively or additionally, the determined planned travel trajectory only deviates at least partially from the learned travel trajectory if at least one obstacle is detected along the learned travel trajectory. For example, if it is already determined at the starting point that an obstacle, such as a garbage can, a pedestrian, and / or a parked vehicle, is detected along the learned travel trajectory, it can be provided that the travel trajectory is at least partially guided around this obstacle. For this purpose, the corresponding planned travel trajectory is automatically determined and then followed in follow-up mode.

[0029] Alternatively or additionally, the determined planned driving trajectory can only deviate at least partially from the learned driving trajectory if, when compared with the driving information describing the vehicle's journey, an erroneous localization of the vehicle is detected. This is always the case, for example, when repetitive and thus always identical objects are arranged in the environment. The driving information mentioned here is the same driving information that has already been described.For example, if the vehicle has to drive past several garage entrances or garage doors along the trained driving trajectory, all of which are identical or at least similar in design, the vehicle's localization may experience a jump due to, for example, a sudden mislocalization, meaning the vehicle is not located in front of the garage entrance or garage door in front of which it is actually currently located. Such a shift in the vehicle's position relative to the driving trajectory, which occurs when the localization algorithm is applied, can be concluded that the localization is faulty. Furthermore, this results in a different current location of the vehicle being determined, one that, for example, is not on the trained driving trajectory and / or is not at the starting location or destination.In such a case, the vehicle should initially return to the learned driving trajectory, which is why the planned driving trajectory, which is at least partially different, can be determined and specified. The erroneous localization can be detected, for example, by comparing it with the driving information, because, for example, a steering behavior and / or a distance traveled by the vehicle does not match the location where the vehicle was incorrectly localized. The location can be described by a position and an orientation.

[0030] Ultimately, various initial situations are possible in which the method according to the invention is particularly useful, since in these the planned driving trajectory deviates at least partially from the learned driving trajectory and therefore the described checking taking into account the second tolerance range is particularly useful.

[0031] The check is intended to take place before the driver assistance system specifies at least assisted longitudinal and / or lateral guidance of the vehicle. It can therefore be specified that the procedure is carried out when the vehicle is not yet moving along the determined planned travel trajectory. The procedure is then started right at the beginning, for example, immediately after the manual or automatic activation of the tracking mode. This clearly demonstrates the desired advantage that the procedure does not suddenly detect the localization problem while operating in tracking mode, but rather that the planned travel trajectory is checked in advance with regard to the localization options along the planned travel trajectory.

[0032] Furthermore, one embodiment provides that the check is always performed for the entire planned travel trajectory from a current location of the vehicle to a destination. The current location can be the starting location of the learned travel trajectory or can be distant from it. Preferably, the entire route to the destination, i.e., the entire duration during which the tracking mode is expected to be activated, is checked in advance. If this is not possible, for example, due to poor visibility conditions, a subsection of the entire planned travel trajectory can be checked first. The travel trajectory can thus be divided into partial travel trajectories.

[0033] Furthermore, one embodiment provides for the driver assistance system to be a trained parking assistant. Alternatively or additionally, the driver assistance system is a reversing assistant. The reversing assistant is designed to learn, for example, a route traveled forwards from the starting point to the destination as a driving trajectory in learning mode. In follow-up mode, the driving trajectory can be driven backwards, i.e. the vehicle now drives backwards from the destination to the starting point. The original destination therefore becomes the new starting point and the original starting point becomes the new destination. Deviations from the learned driving trajectory arise here, for example, in curves due to different movement radii of the vehicle when driving forwards and backwards. For this reason, a planned driving trajectory is also determined here, which can deviate from the trained driving trajectory at least in the area of ​​the curve.The method according to the invention is preferably intended for these types of trained driver assistance systems. Alternative or additional driver assistance systems are possible, which are ultimately based on the principle that the driving trajectory is learned in the learning mode and the learned driving trajectory is followed in the follow-up mode.

[0034] A further aspect of the invention relates to a vehicle configured to carry out the method described above. The vehicle is preferably a motor vehicle, in particular a passenger car, a truck, a bus, a motorcycle, and / or a moped.

[0035] An additional aspect of the invention relates to a control device for the vehicle. The control device is configured to carry out the method described above. The control device carries out the method. If environmental information is required for a method step, it is assumed that this information was previously detected by a sensor device of the vehicle and provided to the control device, so that the control device can now carry out the method based on the environmental information. The control device is an electronic computing device and thus an electronic control unit (Electronic Control Unit, ECU) of the vehicle.

[0036] The control device comprises, for example, a processor device. This can comprise at least one microprocessor, microcontroller, FPGA (Field Programmable Gate Array), and / or DSP (Digital Signal Processor). Furthermore, it can comprise program code, which can alternatively be referred to as a computer program product. The program code can be stored in a data memory of the processor device.

[0037] Another aspect of the invention relates to a computer program product. The computer program product is a computer program. The computer program product comprises instructions which, when the program is executed by a computer, such as, for example, by the control devices of the vehicle, cause the computer to carry out the method according to the invention. The exemplary embodiments described in connection with the method according to the invention, each individually and in combination with one another, apply accordingly, where applicable, to the motor vehicle according to the invention, the control device according to the invention, and the computer program product according to the invention. The invention encompasses combinations of the described exemplary embodiments.

[0038] Showing:

[0039] Fig. 1 A schematic representation of a vehicle with a driver assistance system; and

[0040] Fig. 2 shows a schematic representation of a signal flow graph of a method for operating a driver assistance system in a vehicle.

[0041] In the figures, functionally identical elements are provided with the same reference numerals.

[0042] Fig. 1 shows a vehicle 1 with a driver assistance system 2. The driver assistance system 2 can be a trained parking assistant and / or a reversing assistant. The vehicle 1 can have a control device 3 that is designed to operate the driver assistance system 2. The driver assistance system 2 can be designed to at least assist in specifying, i.e., controlling, a longitudinal and / or lateral guidance of the vehicle 1. It can therefore be designed to accelerate, decelerate, and / or steer the vehicle 1. The driver assistance system 2 is preferably designed to fully automatically control the longitudinal and / or lateral guidance of the vehicle 1. The vehicle 1 can also have a sensor device 4. This can comprise at least one camera, a radar device, and / or a LiDAR device.

[0043] Furthermore, Fig. 1 shows a starting location 6 at which the vehicle 1 is currently located. The starting location 6 here is also a starting location 6 of a learned driving trajectory 5. The driver assistance system 2 is designed to learn the driving trajectory 5 in a learning mode and to follow the learned driving trajectory 5 in a following mode 26 (see reference numeral 26 in Fig. 2). For this purpose, it travels in the following mode 26 from the starting location 6 to a destination 7, which in this case is located in a parking lot 8. The learned driving trajectory 5 ends at the destination 7. For the learned driving trajectory 5, an area 10 can be specified, within which the learned driving trajectory 5 and any planned driving trajectories 12, 13 deviating from it must lie in order to be permitted by the driver assistance system 2.Here, two planned travel trajectories 12, 13 are shown as examples, which deviate at least locally from the learned travel trajectory 5 in order to avoid an obstacle 9. The obstacle 9 is, for example, a garbage can, which is positioned at least temporarily between the starting location 6 and the destination 7. Another obstacle 9 or multiple obstacles 9 are possible. It is clear here that the respective planned travel trajectory 12, 13 deviates from the learned travel trajectory 5 at least for a certain distance 14 due to the obstacle 9. The planned travel trajectory 12 deviates only within the area 10 in order to avoid the obstacle 9, whereas the planned travel trajectory 13 partially leaves the area 10.

[0044] The three travel trajectories 5, 12, and 13 depicted in Fig. 1 are purely exemplary. Alternative shapes, numbers of travel trajectories 5, 12, and 13, and / or other properties of the respective travel trajectories 5, 12, and 13 are possible.

[0045] Fig. 2 shows a method for operating the driver assistance system 2 in the vehicle 1. In a method step S1, the vehicle 1 is localized relative to the learned driving trajectory 5 by applying a localization algorithm 20 and environmental information 21 that describes an environment of the vehicle 1. This determines, for example, localization information 22 that indicates, for example, where the vehicle 1 is located in the environment and thus, for example, relative to the learned driving trajectory 5 and / or the obstacle 9 and / or to at least one object in the environment of the vehicle 1. The environmental information 21 can be determined by the sensor device 4 of the vehicle 1 and provided to the control device 3. The control device 3 can then carry out method step S1 and the following method steps S2 to S5.

[0046] In a method step S2, the planned travel trajectory 12, 13 is determined by applying a planning criterion 23 to the learned travel trajectory 5 and the environmental information 21. A deviation between the determined planned travel trajectory 12, 13 and the learned travel trajectory 5 lies within a first tolerance range 24 specified for the planning criterion 23. Here, for example, it is checked whether the planned travel trajectory 12, 13 lies within the area 10 if the area 10 sketched in Fig. 1 is specified, for example, by the first tolerance range 24. Here, for example, it is determined that the planned travel trajectory 13 lies outside the first tolerance range 24, and thus it can be discarded, for example, and only the planned travel trajectory 12 can be determined.

[0047] In a method step S3, a check is performed to determine whether the deviation between the determined planned travel trajectory 12, 13 and the learned travel trajectory 5 also lies within a second tolerance range 25 specified for localization. The first tolerance range 24 and the second tolerance range 25 deviate at least partially from one another. For example, a different specification regarding the area 10 can be provided here, so that, for example, the determined planned travel trajectory 12, 13 does or does not exhibit a deviation that lies within the respective tolerance range 24, 25.

[0048] In general, the respective tolerance range 24, 25 can specify the area 10 around the learned travel trajectory 5 within which the planned travel trajectory 12, 13 may run. Alternatively or additionally, the respective tolerance range 24, 25 can specify a permitted maximum change in the orientation of the vehicle 1 according to the planned travel trajectory 12, 13, compared to the orientation of the vehicle 1 according to the learned travel trajectory 5. Alternatively or additionally, the respective tolerance range 24, 25 can specify a permitted maximum distance 14 for which the planned travel trajectory 12, 13 may deviate from the learned travel trajectory 5.

[0049] Here, for example, the relatively long distance 14 may be too long for the planned travel trajectory 13 and thus greater than the permitted maximum distance 14, so that this planned travel trajectory 13, even if it were to lie within the area 10, would then have, for example, a deviation between the determined planned travel trajectory 13 and the learned travel trajectory 5 that lies outside the second tolerance range 25. In this example, a method step S5 can be carried out in which an alternative planned travel trajectory 12, 13 is determined and checked, i.e., for example, method steps S2 and S3 are carried out again and / or method step S1 is already carried out again. Alternatively or additionally, the follow-up mode 26 can at least be paused, in particular deactivated.

[0050] However, if it is determined during the checking in method step S3 that the deviation between the planned travel trajectory 12, 13 and the learned travel trajectory 5 lies within the second tolerance range 25, the following of the planned travel trajectory 12, 13 in the following mode is permitted in a method step S4.

[0051] It can be provided that the second tolerance range 25 is predetermined at least partially depending on the applied localization algorithm 20. In general, the respective tolerance range 24, 25 can be at least partially dependent on a number and / or arrangement of the sensor device 4, in particular the plurality of sensor devices 4, in the vehicle 1 and / or on a quality of the environmental information 21. The environmental information 21 can include data from the camera, the radar device, or the LiDAR device as the sensor device 4.

[0052] The check in method step S3 preferably takes place before the driver assistance system 2 specifies the at least assisted longitudinal and / or lateral guidance of the vehicle 1. Preferably, method step S3 always takes place for the entire planned travel trajectory 12, 13 from the current location and thus from the starting location 6 of the vehicle 1 to the destination 7. Thus, the entire planned travel trajectory 12, 13 can be checked in advance.

[0053] If no obstacle 9 is present, for example, the planned travel trajectory 12, 13 can exactly follow the learned travel trajectory 5. However, there are various situations in which the determined planned travel trajectory 12, 13 can deviate at least partially from the learned travel trajectory 5. These situations occur, for example, when the vehicle 1 is manually moved away from the learned travel trajectory 5 and / or at least the obstacle 9 is detected along the learned travel trajectory 5 and / or an erroneous localization of the vehicle 1 is detected upon comparison with driving information describing a journey of the vehicle 1. The latter erroneous localization can occur, for example, by shifting the localization relative to an actual location of the vehicle 1 in the environment.

[0054] Overall, the examples demonstrate how journey planning and thus the determination of the planned journey trajectory 12, 13 can take into account the localization availability and localization capabilities of the driver assistance system 2. To avoid a deterioration in the performance (accuracy) of the localization or the aborting of an active driving maneuver along the journey trajectory 5, 12, 13, the driver assistance system 2 must plan the journey trajectory 12, 13 such that the future positions of the vehicle 1 do not lead to a deterioration in the performance of the localization algorithm 20. For example, the localization algorithm 20 can only localize with good quality to the positions of the previously recorded trajectory (learned journey trajectory 5) below certain distance and orientation thresholds.This naturally depends on the localization algorithm 20 used and the number of sensor devices 4 used, so those planned driving trajectories 12, 13 that exceed the deviation thresholds (second tolerance range 25) should be avoided, or, if this is not possible, those driving trajectories 12, 13 should be planned for which the deviation does not exceed one percent. Alternatively, it can be specified that only those driving trajectories 12, 13 are planned for which the deviation thresholds of the localization algorithm 20 are exceeded only for a specific traveled route 14 (since the position of the vehicle 1 can be tracked by the odometry (driving information) of the vehicle 1 during the degraded localization performance).One implementation could be that during a trajectory search with a*, trajectories that deviate too greatly from the planned driving trajectory 12, 13 from the learned driving trajectory 5 are penalized, so that the planned driving trajectories 12, 13 that exceed the distance above the deviation threshold, i.e., lie outside the second tolerance range 25, are discarded.

Claims

Patent claims 1. A method for operating a driver assistance system (2) in a vehicle (1), wherein the driver assistance system (2) is designed to learn a driving trajectory in a learning mode and to follow the learned driving trajectory (5) in a tracking mode (26), wherein the method in the tracking mode (26) comprises: - Localizing (S1) the vehicle (1) relative to the learned driving trajectory (5) by applying a localization algorithm (20) to environmental information (21) describing an environment of the vehicle (1); - determining (S2) a planned travel trajectory (12, 13) by applying a planning criterion (23) to the learned travel trajectory (5) and the environmental information (21), wherein a deviation between the determined planned travel trajectory (12, 13) and the learned travel trajectory (5) lies within a first tolerance range (24) predetermined for the planning criterion (23); - checking (S3) whether the deviation between the determined planned travel trajectory (12, 13) and the learned travel trajectory (5) is also within a second tolerance range (25) specified for localization, which deviates at least partially from the first tolerance range (24); - only if this is the case, allowing (S4) the following of the planned travel trajectory (12, 13) in the following mode (26).

2. The method according to claim 1, wherein if the deviation between the determined planned travel trajectory (12, 13) and the learned travel trajectory (5) lies outside the second tolerance range (25), an alternative planned travel trajectory (12, 13) is determined and checked and / or the follow-up mode (26) is at least paused, in particular deactivated (S5).

3. Method according to one of the preceding claims, wherein the respective tolerance range (24, 25) specifies an area (10) around the learned travel trajectory (5) within which the planned travel trajectory (12, 13) may run.

4. Method according to one of the preceding claims, wherein the respective tolerance range (24, 25) specifies a permitted maximum change in an orientation of the vehicle (1) according to the planned travel trajectory (12, 13) compared to the orientation of the vehicle (1) according to the learned travel trajectory (5).

5. Method according to one of the preceding claims, wherein the respective tolerance range (24, 25) specifies a permitted maximum distance (14) for which the planned travel trajectory (12, 13) may deviate from the learned travel trajectory (5).

6. Method according to one of the preceding claims, wherein the second tolerance range (25) is predetermined at least partially as a function of the localization algorithm (20) used.

7. Method according to one of the preceding claims, wherein the respective tolerance range (24, 25) is predetermined at least partially as a function of a number and / or arrangement of sensor devices (4) in the vehicle (1) which detect and provide the environmental information (21), and / or a quality of the environmental information (21).

8. The method according to claim 7, wherein the environmental information (21) comprises data from at least one camera, a radar device and / or a LiDAR device as the sensor device (4).

9. Method according to one of the preceding claims, wherein the determined planned travel trajectory (12, 13) deviates at least partially from the learned travel trajectory (5) only if: - the vehicle (1) is manually moved away from the learned driving trajectory (5); and / or - at least one obstacle (9) is detected along the learned driving trajectory (5); and / or - when comparing with driving information describing a journey of the vehicle (1), an erroneous localization of the vehicle (1) is detected.

10. Method according to one of the preceding claims, wherein the checking is carried out before the driver assistance system (2) specifies at least assisted longitudinal and / or lateral guidance of the vehicle (1).

11. Method according to one of the preceding claims, wherein the checking is always carried out for the entire planned travel trajectory (12, 13) from a current location of the vehicle (1) to a destination (7).

12. Method according to one of the preceding claims, wherein the driver assistance system (2) is a trained parking assistant and / or a reversing assistant.

13. Vehicle (1) designed to carry out a method according to one of the preceding claims.

14. Control device (3) for a vehicle (1), wherein the control device (3) is designed to carry out a method according to one of claims 1 to 12.

15. A computer program product comprising instructions which, when executed by a computer, cause the computer to perform a method according to any one of claims 1 to 12.

Images