Method for providing a future trajectory for an at least partially assisted motor vehicle and assistance system

Abstract

Method for providing a future trajectory (4) for an at least partially assisted motor vehicle (5) by means of an assistance system (1), comprising the steps: - detecting a current position (P) of the motor vehicle (5) by means of a detection device (6) of the motor vehicle (5); - comparing the current position (P) with at least one position (9) stored in a memory device (3) of the assistance system (1), wherein at least one historically traveled trajectory (7, 8) of at least one further motor vehicle is assigned to the stored position (9); and - If the current position (P) matches the stored position (9), providing at least one partial trajectory of the trajectory (7, 8) already traveled historically as a trajectory (4) that can be traveled in the future, wherein the at least one trajectory (7, 8) already traveled historically is provided on an electronic computing device (2) of the assistance system (1) external to the motor vehicle, wherein a respective point in time from the respective further trajectories (7, 8) traveled historically is taken into account when determining the approximated trajectory and wherein older trajectories (7, 8) traveled historically are given less weighting than new ones.

Description

[0001] The invention relates to a method for providing a future trajectory for an at least partially assisted motor vehicle by means of an assistance system according to the applicable patent claim 1. Furthermore, the invention relates to a computer program product and an assistance system.

[0002] So-called trained parking is already known from the state of the art, whereby a trajectory to a specified end position can be learned. This learning process is also referred to as teach-in. The vehicle later automatically drives the learned route, which is also known as redrive mode. The vehicle user can be inside or outside the vehicle, can monitor the assisted function, or the vehicle can drive autonomously without driver supervision.

[0003] However, the driver may not be able or willing to learn the trajectory for trained parking. The driver may not be sufficiently familiar with the vehicle's functions, or the training process may be too time-consuming or complicated. As a result, the driver repeatedly drives manually, even though the route could be completed using an assistance function for parking or the parking process itself.

[0004] DE 10 2013 015 348 A1 relates to a method for operating a vehicle, in particular for driving the vehicle to a parking space in a parking zone that is not visible / far from the street, in which environmental data of the vehicle are recorded, wherein when driving to a parking space in the parking zone it is identified whether this is a home parking space or the parking zone is a home parking zone, and when the home parking space or home parking zone is identified and the vehicle approaches the identified home parking space or the identified home parking zone, recorded environmental data or driving data are stored or updated, wherein in a learning mode, based on the environmental data or driving data, several trajectories for the at least one home parking space of the at least one home parking zone are determined and stored,and wherein, in an operating mode, when approaching the at least one home parking space of the at least one home parking zone, possible drivable trajectories are output for selection and activation, or if only one of the determined trajectories is drivable, this is automatically activated.

[0005] DE 10 2014 018 189 A1 relates to a method for operating a vehicle in at least one parking situation. At least one predefined vehicle-specific or parking situation-specific function is assigned to at least one predefined parking situation of the vehicle. It is determined whether the vehicle is in the at least one predefined parking situation. The predefined vehicle-specific or parking situation-specific function assigned to this at least one predefined parking situation is offered to a vehicle user for activation or is automatically activated when the vehicle is in the at least one predefined parking situation.

[0006] DE 10 2017 203 983 A1 discloses a method for operating a motor vehicle when carrying out a manoeuvring operation in a manoeuvring environment that can be described by position information, wherein, depending on the current position information of the motor vehicle, at least one piece of maneuver information assigned to the manoeuvring environment and describing a driving maneuver for the manoeuvring operation is selected and used to carry out the manoeuvring operation, wherein the selected maneuver information is received by a server device external to the motor vehicle via a communication connection.

[0007] The object of the present invention is to provide a method and an assistance system by means of which a user of the motor vehicle can be provided with a trajectory for at least partially assisted operation.

[0008] This object is achieved by a method and an assistance system according to the independent patent claims. Advantageous embodiments are specified in the subclaims.

[0009] One aspect of the invention relates to a method for providing a future trajectory for an at least partially assisted motor vehicle by means of an assistance system, comprising detecting a current position of the motor vehicle by means of a detection device of the motor vehicle. The current position is compared with at least one position stored in a memory device of the assistance system, wherein at least one previously historically traveled trajectory of another motor vehicle is assigned to the stored position. If the current position matches the stored position, at least a partial trajectory of the previously historically traveled trajectory is provided as a future trajectory.

[0010] This will improve the future trajectory for the user of the motor vehicle.

[0011] In particular, a method is proposed in which trajectories for automated parking and maneuvering, such as trained parking or a reversing assistant, can be independently learned by recording and evaluating trajectories or traveled distances. In this process, identical sections and recurring route sections, such as a yard entrance, are recognized, and based on the repeated journeys, driver information is generated in various forms, including a retrievable trajectory. For the detected route sections, the trajectory is offered to the user using a suitable function, such as trained parking.

[0012] In particular, artificial intelligence, statistical models, feature assignments through, for example, feature-landmark matching, Monte Carlo association, Markov chain or other machine learning methods can be used to learn the parking or maneuvering processes.

[0013] In particular, a redrive function based on other motor vehicles is proposed. The motor vehicle itself recognizes recurring characteristics and behavior in the background, which are also associated with the vehicle position, in particular globally or map-relatively, from previously performed parking and maneuvering maneuvers of other motor vehicles with the same or similar destination, and retrieves the resulting redrive trajectories from the storage device. The storage device can preferably be provided, for example, on an electronic computing device external to the vehicle. The electronic computing device external to the vehicle can also be referred to as a backend. Alternatively, the electronic computing device or storage device can also be provided internally to the vehicle, whereby the corresponding trajectory can also be obtained, for example, via C2X communication (car-to-infrastructure).

[0014] The future trajectory is then generated by the storage device and made available to the user. The future trajectory is generated based on the trajectory already traveled in the past. Entire trajectories or just partial trajectories can be used or determined.

[0015] Furthermore, the at least one trajectory already traveled historically is provided on an electronic computing device of the assistance system external to the vehicle. In particular, for example, the current position of the motor vehicle can then be transmitted to the electronic computing device external to the vehicle, wherein this position is compared with the position already assigned to a trajectory. At least the partial trajectory is then provided, in particular transmitted, to the motor vehicle as a trajectory that can be followed in the future. This saves storage capacity, and in particular, it is possible to check at many locations up-to-date whether a corresponding trajectory exists.

[0016] It is also provided that a respective point in time from the respective other historically driven trajectories is taken into account when determining the approximated trajectory. In particular, it can be provided, for example, that older historically driven trajectories are given less weighting than new ones. In particular, this can also ensure that changes within the environment are taken into account. For example, new signs or objects may have been placed along the trajectory, with the more recent historically driven trajectories bypassing this sign or object. In this way, a trajectory that can currently be followed in the future can be made available to the user. The at least one trajectory that has already been followed historically is made available on an electronic computing device of the assistance system external to the vehicle.In particular, for example, the current position of the motor vehicle can then be transmitted to the electronic computing device external to the vehicle, whereby this position is compared with the position already assigned to a trajectory. At least the partial trajectory is then provided, in particular transmitted, to the motor vehicle as a trajectory that can be followed in the future. This saves storage capacity and, in particular, allows for a current check at many locations to determine whether a corresponding trajectory exists.

[0017] According to a further advantageous embodiment, upon repeated approach to a possible starting position, the recording function is automatically offered if it is determined that no historical trajectories are yet available for this possible starting position. In particular, this embodiment eliminates the need for an explicit teach-in, so that the user can perform the parking or maneuvering process more conveniently by saving the recorded trajectory as a historical trajectory and thus making it available for future journeys. The teach-in option is offered after detecting repeated approach to the possible starting position, and the trajectory, or partial trajectory, is then recorded.Alternatively, the recording can also be carried out automatically as soon as the repeated approach is detected, and after the teach-in is completed, the driver can be offered the opportunity to save the recorded route for future situations.

[0018] According to an advantageous embodiment, the at least one previously traveled trajectory is generated based on a plurality of additional motor vehicles. In particular, the previously traveled trajectory can thus be generated based on swarm data from a plurality of motor vehicles, which have been in a similar position or location, in particular. This can be achieved, in particular, by merging the trajectories and, for example, by approximating them.

[0019] It is further advantageous if, based on further historically driven trajectories of the plurality of motor vehicles, an approximated trajectory is determined as at least the partial trajectory and provided to the motor vehicle. In particular, since not all motor vehicles follow exactly the same trajectory, it is therefore advantageous if the trajectory is only approximated based on the historically driven trajectories. In particular, this allows corresponding outliers to be identified during the learning process or the generation process and eliminated from the trajectory used. For example, a one-time braking or bypassing of a temporary obstacle can be removed from the redrive trajectory.

[0020] It has also proven advantageous if the at least one partial trajectory is output to a user of the motor vehicle for selection on an output device. In particular, the user can have the corresponding trajectory or a plurality of trajectories output to them, for example, on a display device as the output device. The user can then choose whether they want to follow the respective trajectory. This allows the user to intuitively select the trajectory that can be followed in the future. Inconvenient learning of a corresponding trajectory is no longer necessary in this case.

[0021] It has also proven advantageous to evaluate the previously traveled trajectory using machine learning on the electronic computing device. Artificial intelligence, such as a neural network, can be used for this purpose. Furthermore, statistical models, feature assignment through feature-landmark matching, Monte Carlo association, or even a Markov chain can be used. This provides the user with an improved future trajectory.

[0022] In a further advantageous embodiment, the future traversable trajectory is provided for an assisted maneuvering of the motor vehicle and / or for an assisted parking of the motor vehicle. This is particularly possible for recurring positions. Thus, for example, at a drive-thru that is visited frequently, a correspondingly assisted maneuvering process can be provided. Furthermore, a parking process, for example, in a home zone, can also be provided.

[0023] The method presented is a computer-implemented method. A further aspect of the invention relates to a computer program product with program code means that, when an electronic computing device executes the program code means, cause it to perform a method according to the preceding aspect.

[0024] For this purpose, the electronic computing device comprises, in particular, electronic components, processors and, for example, integrated circuits in order to be able to carry out the corresponding process.

[0025] Yet another aspect of the invention relates to an assistance system for providing a future trajectory for an at least partially assisted motor vehicle, comprising at least one storage device. The assistance system is designed to carry out a method according to the preceding aspect. In particular, the method is carried out by means of an assistance system.

[0026] The assistance system is preferably implemented externally of the vehicle. Alternatively, however, the assistance system can also be implemented internally. Furthermore, some features of the assistance system can be implemented externally of the vehicle, while others can be implemented internally.

[0027] Advantageous embodiments of the method are to be regarded as advantageous embodiments of the computer program product and the assistance system. The assistance system has tangible features that enable implementation of the method or an advantageous embodiment thereof.

[0028] The invention also includes further developments of the assistance system according to the invention that have features already described in connection with the further developments of the method according to the invention. For this reason, the corresponding further developments of the assistance system according to the invention are not described again here.

[0029] The invention also includes the combination of the features of the described embodiments.

[0030] Exemplary embodiments of the invention are described below. Shown are: Fig. 1 a schematic plan view of an embodiment of an assistance system; and Fig. 2 a schematic flow diagram according to an embodiment of the method.

[0031] The exemplary embodiments explained below are preferred exemplary embodiments of the invention. In the exemplary embodiments, the described components each represent individual, independently considered features of the invention, which also further develop the invention independently of one another and are thus also to be considered as components of the invention, either individually or in a combination other than that shown. Furthermore, the described exemplary embodiments can also be supplemented by further features of the invention already described.

[0032] In the figures, functionally identical elements are provided with the same reference numerals.

[0033] Fig. 1 shows a schematic plan view of an embodiment of an assistance system 1. In the present case, the assistance system 1 comprises an electronic computing device 2 external to the motor vehicle. The electronic computing device 2 external to the motor vehicle further comprises a memory device 3. The assistance system 1 is designed to provide a future trajectory 4 for a motor vehicle 5 that is operated at least partially with assistance.

[0034] In the method for providing the future trajectory 4 for the at least partially assisted motor vehicle 5, a current position P of the motor vehicle 5 is detected by means of a detection device 6 of the motor vehicle 5. In this case, the current position P is transmitted in particular to the electronic computing device 2 external to the motor vehicle. The current position P is then compared with at least one position 9 stored in the memory device 3, wherein at least one trajectory 7, 8 of another motor vehicle that has already been driven historically is assigned to the stored position 9. If the current position P matches the stored position 9, at least one partial trajectory of the trajectories 7, 8 that have already been driven historically is provided as the future trajectory 4.

[0035] In particular, it is provided that the at least one historically driven trajectory 7, 8 is generated based on a plurality of additional motor vehicles. Furthermore, based on the respective additional historically driven trajectories 7, 8 of the plurality of motor vehicles, an approximated trajectory is determined as at least the partial trajectory and provided to the motor vehicle 5.

[0036] In the present case, two trajectories 7, 8 that have already been driven historically are shown. In particular, a first trajectory 7 that has already been driven historically by, for example, a first additional motor vehicle and a second trajectory 8 that has already been driven historically by a second additional motor vehicle are shown.

[0037] Furthermore, it can be provided that a respective point in time from the respective further historically traveled trajectories 7, 8 is taken into account when determining the approximated trajectory.

[0038] It can further be provided that the at least one partial trajectory is output to a user of the motor vehicle 5 for selection on an output device 10.

[0039] Furthermore, it can be provided that the trajectories 7, 8 already traveled historically are evaluated by means of machine learning of the electronic computing device 2.

[0040] Furthermore, it can be provided that the trajectory 4 that can be followed in the future is provided for an assisted maneuvering process of the motor vehicle 5 or, as shown here, for an assisted parking process of the motor vehicle 5, for example in a parking space 11.

[0041] In particular, a method for the independent learning of trajectories for automatic parking or maneuvering is proposed, in which the previously traveled trajectories 7, 8 are recorded and evaluated. In this process, identical sections and recurring route sections, such as a farmyard entrance, are recognized, and based on the repeated journeys, driver information is generated in various forms up to the future trajectory 4. For the identified route sections, the future trajectory 4 is offered to the user using a suitable function, such as trained parking.

[0042] In the background, the motor vehicle 5 recognizes recurring features and behavior, also together with the position P, in particular globally or map-relatively, from previously performed parking and maneuvering maneuvers of other motor vehicles with the same or similar destinations and retrieves the resulting future drivable trajectory 4 from the electronic computing device 2 external to the motor vehicle in this exemplary embodiment. The so-called redrive of the trajectory generated and made available by the electronic computing device 2 external to the motor vehicle is offered to the user. Upon repeated approach to a possible starting position, the function is offered automatically. This eliminates the need for an explicit teach-in for the user of the motor vehicle.

[0043] The function can be executed within the motor vehicle 5, so that potentially personal data does not leave the motor vehicle 5. Alternatively or additionally, depending on the customer's wishes, the data, such as positions, trajectories, times of recording, etc., can also be processed on the electronic computing device 2 external to the motor vehicle or made directly usable with other motor vehicles using C2X (car-to-infrastructure) methods.

[0044] Through the various learning processes, outliers are identified and eliminated from the trajectory used. For example, a single braking or circumvention of a temporary obstacle can be removed from the redrive trajectory.

[0045] The information that the motor vehicle 5 or, in this case, the electronic computing device 2 external to the motor vehicle can provide or use includes, among other things, map data, environmental data or trajectories of other motor vehicles or other road users as well as landmarks and features for localization.

[0046] In particular, in this case, an entire parking process can be assumed. Alternatively or additionally, the usage can be further developed so that only sections of trajectories, i.e., partial trajectories, need to be compared and recognized. Accordingly, not only a complete redrive trajectory can be used, but also sections of other trajectories.

[0047] In particular, it can further be provided that the assistance system 1 recognizes the repeated driving process and informs the user of the motor vehicle 5 depending on the situation, for example via the output device 10, wherein a corresponding teach-in function can then be provided. The motor vehicle 5 records recurring characteristics and behavior in the background, possibly together with the position P, in order to recognize frequently used, recurring parking and maneuvering processes. If a repeated process is recognized, the user of the motor vehicle 5 is informed of the availability of the function, and the teaching of an optimal trajectory can be offered. The recognition and driver interaction can occur both during the start-up process and upon completion of the process. However, explicit teach-in by the driver is necessary in this case.The trajectory learned during teach-in can be enhanced using previously performed maneuvers and / or existing backend data. The learned trajectory can then be used for redrive.

[0048] Alternatively, the assistance system 1 can detect the repeated driving process and offer the user the option of saving the trajectory generated in the background. The motor vehicle 5 records recurring characteristics and behavior in the background, possibly together with the position P, in order to detect frequently used, repetitive parking and maneuvering processes. If a repeated process is detected sufficiently well for a redrive, the driver is informed of the availability of the function and the option of saving the generated trajectory is offered. The detection and driver interaction can occur both during the start process and upon completion of the process. The redrive of the learned trajectory and / or the trajectory enriched by the backend is offered to the driver, for example, from the output device 10. If the starting position is approached repeatedly, the function is offered automatically.This also eliminates the need for an explicit, manual teach-in by the driver.

[0049] Fig.2 shows a schematic flow diagram according to an embodiment of the method. The method begins in a first step S1. In a second step S2, a search is made for recurring trajectories. In a first alternative, starting from the second step S2, a transition can be made to a third step S3, in which the recurring trajectory is recognized and a check is carried out to determine whether it can be provided with sufficient quality for the redrive. In a fourth step S4, the redrive is offered, and in a fifth step S5, the redrive is carried out. Starting from the second step S2, a transition can be made to a sixth step S6, in which a check is carried out to determine whether a recurring trajectory is recognized. If this is the case, a so-called teach-in can be offered in a seventh step S7.In an eighth step S8, the teach-in is then performed again, and from the eighth step S8, the system can then proceed to the fifth step S5. Alternatively, from the second step S2, the system can proceed to a ninth step S9, where, in the ninth step S9, the start of the trajectory of another motor vehicle matching the system's own destination is detected. In a tenth step S10, the redrive is offered, and from the tenth step S10, the system can again proceed to the fifth step S5 and perform the redrive.

Claims

[1] Method for providing a future trajectory (4) for an at least partially assisted motor vehicle (5) by means of an assistance system (1), comprising the steps: - detecting a current position (P) of the motor vehicle (5) by means of a detection device (6) of the motor vehicle (5); - comparing the current position (P) with at least one position (9) stored in a memory device (3) of the assistance system (1), wherein at least one historically traveled trajectory (7, 8) of at least one further motor vehicle is assigned to the stored position (9); and - If the current position (P) matches the stored position (9), providing at least one partial trajectory of the trajectory (7, 8) already traveled historically as a trajectory (4) that can be traveled in the future, wherein the at least one trajectory (7, 8) already traveled historically is provided on an electronic computing device (2) of the assistance system (1) external to the motor vehicle, wherein a respective point in time from the respective further trajectories (7, 8) traveled historically is taken into account when determining the approximated trajectory and wherein older trajectories (7, 8) traveled historically are given less weighting than new ones. [2] Method according to claim 1, characterized by that the at least one trajectory (7, 8) which has already been driven historically is generated on the basis of a large number of other motor vehicles. [3] Method according to claim 2, characterized bythat on the basis of respective further historically driven trajectories (7, 8) of the plurality of motor vehicles, an approximated trajectory is determined as at least the partial trajectory and is provided to the motor vehicle (5). [4] Method according to one of the preceding claims, characterized by that the at least one partial trajectory is output to a user of the motor vehicle (5) for selection on an output device (10). [5] Method according to one of the preceding claims, characterized by that the trajectory (7, 8) already traveled historically is evaluated by means of machine learning of an electronic computing device (2). [6] Method according to one of the preceding claims, characterized by that the future drivable trajectory (4) is provided for an assisted maneuvering operation of the motor vehicle (5) and / or for an assisted parking operation of the motor vehicle (5). [7] Method according to one of the preceding claims, characterized by that after detecting the current position (P) and determining a match between the current position (P) and at least one previously stored position and the absence of a previously historically traveled trajectory (7, 8) for the current position (P) in the storage device (3) of the assistance system (1), at least one of the steps is carried out: - recording a trajectory driven from the current position (P) and offering to store the recorded trajectory as a historically driven trajectory in the storage device (3) of the assistance system (1), or - Offering to store the trajectory still to be driven, recording the trajectory driven from the current position (P) and storing the recorded trajectory as a historically driven trajectory in the S storage device (3) of the assistance system (1). [8] Computer program product with program code means which cause an electronic computing device (2) to carry out a method according to one of claims 1 to 7 when it executes the program code means. [9] Assistance system (1) for providing a future trajectory (4) for an at least partially assisted motor vehicle (5), with at least one storage device (3), wherein the assistance system (1) is designed to carry out a method according to one of claims 1 to 7.

Images