Improving a trajectory

Abstract

The invention relates to a method (200) for operating a driver assistance system (110), comprising: providing (S205) a trajectory data set (300) containing multiple waypoints (302) on a trajectory (304), the trajectory (304) having a regular distance (306) between successive waypoints (302) in at least part of the trajectory, searching (S206) the trajectory data set (300) for a first section (308) from a first waypoint (302b) to a second waypoint (302c), which two waypoints (302b, 302c) satisfy the condition that a distance (310) of the first waypoint (302b) from the second waypoint (302c) is shorter than the regular distance (306), generating (S209) a new section (316) for the trajectory (304) from the first waypoint (302b) to the second waypoint (302c), and replacing (S215) the first section (308) with the new section (316).

Description

[0001] 2024PF02762

[0002] 1

[0003] Driver assistance system

[0004] The present invention relates to methods for operating a driver assistance system. The invention also relates to a computer program product, a driver assistance system, and a vehicle.

[0005] In modern vehicles, the number of functions / processes to be executed and the amount of resources required for them are constantly increasing. This includes methods for recording and / or tracing trajectories. Therefore, there is a need for methods for recording and / or tracing trajectories that can operate efficiently using available resources such as storage and processing capacity.

[0006] Against this background, one object of the present invention is to provide means for improving a trajectory in a resource-saving manner.

[0007] Accordingly, a method for operating a driver assistance system is proposed. This method comprises the following steps: providing a trajectory dataset containing multiple waypoints of a trajectory, wherein the trajectory has a regular interval between successive waypoints in at least a portion of the trajectory; searching the trajectory dataset for a first segment from a first waypoint to a second waypoint, for which both waypoints have a distance from the first waypoint to the second waypoint that is shorter than the regular interval; generating a new segment for the trajectory from the first waypoint to the second waypoint; and replacing the first segment with the new segment.

[0008] The method allows for the improvement of a previously recorded trajectory. This means that improvements do not need to be made during the recording run itself. Thus, resources required during the recording run (especially storage capacity and processor capacity) are freed up for determining and recording waypoints and optionally 2024PF02762.

[0009] 2

[0010] The system is kept free for recording environmental features and for operating other driver assistance functions.

[0011] The "trajectory dataset" contains at least several waypoints. One can also say that the trajectory dataset defines, specifies, and / or contains information about the multiple waypoints. A waypoint (technically: keyframe) is at least a two-dimensional positional reference.

[0012] The "trajectory" is preferably a trained trajectory. For example, the driver assistance system or another vehicle system is configured to record and save a manually driven trajectory in a training mode. Various sensor signals are recorded that characterize the vehicle's driving state, such as speed, position, steering angle, and the like, as clearly as possible. Sensor signals from the vehicle's environmental sensors may also be recorded, providing, for example, a representation of the vehicle's surroundings, particularly the position of obstacles. By playing back the vehicle's driving state synchronously, i.e., repeating it, the trained trajectory can be replicated. To replicate the predefined trajectory, it is desirable to consider current environmental sensor data.Therefore, the driver assistance system receives a sensor signal indicative of the environment. This signal can be received directly from one or more of the vehicle's environmental sensors, combining multiple sensor signals from different sensors, or the system can receive the signal in a pre-processed state, for example, as a digital map of the environment showing detected obstacles.

[0013] The trajectory, at least in part, has a "regular" spacing between successive waypoints. This usually occurs because, during a recording run, the waypoints are typically recorded at a fixed or predefined interval. One could also say that 2024PF02762

[0014] 3

[0015] At least most waypoints of the provided dataset are evenly distributed and / or evenly spaced. For example, there is a case where immediately consecutive waypoints all have a regular spacing, but due to a loop, at least two non-consecutive waypoints have a shorter distance between them. There is also a case where, in at least one part of the dataset, immediately consecutive waypoints have a regular spacing, but in at least another part, immediately consecutive waypoints have a shorter distance between them, for example, as a result of a stop during the recording process.

[0016] It should be noted that a "rule" implies that there may be exceptions. For example, if the recording journey was paused at a barrier, a pedestrian crossing, or due to other traffic conditions, or if multi-stage maneuvering was performed during the recording journey, the distances may be shorter than the regular interval. The regular interval corresponds, for example, to the length of the vehicle itself and / or the vehicle that recorded the waypoints of the provided trajectory / trajectory data set.

[0017] A segment can be considered a searchable or discoverable segment if the distance between the first waypoint of that segment and the last waypoint of that segment is shorter than the regular segment length. Two waypoints that are closer to each other than the regular segment length define a searchable or discovered segment. If there are one or more waypoints between the first and second waypoints, these are part of the segment.

[0018] The terms "first" section, "first" waypoint, and "second" waypoint are used for naming purposes. For example, the term "first" section is used even though a "second" section will only be used later, to avoid confusion between "section" and "first section" later on. 2024PF02762

[0019] 4

[0020] The procedure includes a "search" step. Here, the trajectory dataset is searched for two waypoints that are closer together than the regular interval. This is a search criterion. It can also be additionally required that a segment is considered a searchable segment if there is at least one other waypoint between the first waypoint of that segment and the last waypoint of that segment.

[0021] The trajectory dataset may contain an elevation indicator, where a shorter distance than the regular interval is only detected if the elevation difference between the respective waypoints does not exceed a threshold. An elevation indicator could be, for example, an elevation value at each waypoint, a vehicle inclination at each waypoint, floor information in the surrounding data, and so on. The threshold is preferably suitable for distinguishing different floors of a parking garage. For example, the threshold could be 1 meter or less.

[0022] It's possible to simply calculate the distance between each waypoint and every other waypoint. This approach is the least error-prone. If any of these calculated distances is shorter than the regular spacing, or preferably shorter than a predefined percentage of the regular spacing, the respective waypoints are identified as the first and second waypoints. The predefined percentage could be, for example, 80%, 90%, or 95% to minimize the number of sections requiring revision.

[0023] The search process may involve virtually overlaying each waypoint with a control contour, and then recognizing an intersection of two control contours as the shorter distance between the first and second waypoints. In other words, if two control contours overlap, this is interpreted and treated as the shorter distance. This can be described as a virtual overlay or virtual control contours. Searching using control contours has the advantage of being able to detect different distances depending on the direction. 2024PF02762

[0024] 5

[0025] It is possible that every intersection of two control contours, optionally taking elevation information into account, is recognized as the shorter distance between the first waypoint and the second waypoint. This behavior can lead to a relatively large number of examined "first segments," with the result that the trajectory undergoes a comparatively large number of—even relatively small—improvements over its entire length.

[0026] It is possible that only an overlap of two control contours by at least a predetermined area is recognized as the shorter distance between the first waypoint and the second waypoint. This behavior is suitable, for example, to limit the effort to relatively coarse deviations from an ideal trajectory.

[0027] The control contour can simply be a circle around each waypoint. Alternatively, the control contour can have an orientation. In other words, it can extend to different distances in different directions. The control contour at the different waypoints is preferably the same, but not necessarily oriented the same way. It is also possible that different control contours are used depending on waypoint properties, for example, a speed-dependent control contour. A preferred control contour corresponds to an outline of the vehicle and / or the recording vehicle of the trajectory. This outline can be understood, for example, as a vertical projection of the vehicle or its body onto a surface.By choosing a control contour that is at least similar to the vehicle, it is possible, for example, to distinguish between driving in adjacent lanes and driving on or crossing the same lane.

[0028] The method is designed to find trajectory segments where the recorded journey has, for example, loops (e.g., caused by faulty pathfinding, parking search, or evasive maneuvers), spikes, or tight radii. However, it can happen that an initial segment found is caused by external influences or 2024PF02762

[0029] 6

[0030] It simply cannot be "improved." Therefore, the procedure may include the following before replacement: Checking whether the newly created section is drivable. This check could, for example, be a collision check based on detected environmental features. These environmental features could be, for example, currently detected environmental features and / or environmental features stored in the trajectory data set from the recorded journey and / or an interim follow-up journey. This check could, for example, be a drivability check based on detected surfaces, so that, for example, driving over a flowerbed or pond is avoided. This check could, for example, be a traffic rule check based on detected traffic control devices. Traffic control devices could be, for example, road markings, signs, traffic lights, and the like.This avoids making an obviously undesirable change to the trajectory.

[0031] There are many known algorithms for generating a trajectory or a new segment of a trajectory between two waypoints. This procedure is not limited to any one specific algorithm. The new segment is preferably generated to connect smoothly to the first and second waypoints. For example, a Bézier curve might be calculated between the first and second waypoints. Alternatively, a defined library of pre-calculated maneuvers, such as line maneuvers and / or circle maneuvers, might be provided to construct the new segment. It is preferred that the new segment calculates the straightest possible connection between the first and second waypoints. It is also preferred that the new segment runs as close as possible to, and preferably within, a defined path of the recorded trajectory.It is preferred to first calculate a trajectory path in the new section, and then to calculate new waypoints as close as possible to regular intervals along this path.

[0032] It may be that the procedure, before the replacement, generates at least one further new segment a) from a waypoint directly before the first waypoint to the second waypoint.2024PF02762

[0033] 7

[0034] The process includes a waypoint, a new segment (b) from the waypoint immediately before the first waypoint to a waypoint immediately after the second waypoint, and / or a new segment (c) from the first waypoint to the waypoint immediately after the second waypoint. Then, one of the new segments is selected. For clarity, the "new segments" must include at least one of segments (a) through (c), and in any case, the newly generated segment from the first waypoint to the second waypoint. This allows for the testing and achievement of a more comprehensive trajectory improvement. The selection is preferably based on at least one smoothness factor of the respective trajectory.

[0035] The term "smoothness" refers to a mathematical criterion. Smoothness can also be described as a derivative or a kink measure. For example, one can require a maximum distance from a virtual straight line through the previous two waypoints for each waypoint after the third. Similarly, for each triplet of three consecutive waypoints, one can calculate the radius of a circle intersecting all three waypoints and require that this radius change by no more than a specified value / percentage between any two consecutive triplets (i.e., two matching waypoints).

[0036] It is possible that the creation of at least one further section, and preferably the selection of one of the new sections, can be repeated. For example, after selecting one of sections a) to c), its start and end waypoints can be treated as the first and second waypoints. This allows the trajectory improvement to be extended as needed. This can be described as recursive repetition. "Recursive" in this context means, in particular, that starting from the selected section, at least one further section is created whose first waypoint is located directly before the first waypoint of the previously selected section and / or whose last waypoint is located directly after the last waypoint of the previously selected section. 2024PF02762

[0037] 8

[0038] When creating a new section, the recording vehicle's direction of travel may be maintained at both the first and second waypoints. For example, a change between two trains (vehicle trains) can be identified by the fact that, starting from a given waypoint, the preceding and subsequent waypoints lie in the same direction. This allows, for instance, a driver's request to reverse park to be followed, perhaps to ensure a charging point is located next to a wallbox.

[0039] The procedure may include triggering the execution of the remaining steps. For example, the subsequent steps of the procedure may be triggered at regular intervals. This can be described as cyclic triggering or cyclic execution of the procedure. This ensures that all trajectory data sets are regularly optimized.

[0040] The subsequent steps of this process may be triggered by the end of a trajectory data recording. Thus, the trajectory is improved as needed soon after or immediately after the recording ends. Alternatively, the process may wait until the vehicle is switched off. This allows the effort required to improve the trajectory to be deferred to a suitable timeframe.

[0041] It is possible that the subsequent steps of this procedure are executed before a trajectory is followed. For example, the subsequent steps of the procedure are triggered by approaching at least one of the waypoints of the trajectory dataset, in particular by approaching a starting waypoint of the trajectory. One could also say that the steps are triggered by preparing for the trajectory and / or initiating the trajectory and / or starting the trajectory. Thus, one can, for example, limit the effort for this procedure to those trajectory datasets that are at least likely to be followed. 2024PF02762

[0042] 9

[0043] It is also possible that the subsequent steps of this process are triggered by the end of a follow-through along the trajectory. Thus, the trajectory can be improved after a follow-through based on updated / different environmental sensor signals.

[0044] The procedure need not be limited to one of the aforementioned trigger schemes, but rather it can be prepared for situation-dependent use in any of these and / or other cases.

[0045] Furthermore, a computer program product is proposed that includes commands which, when executed by a computer, cause it to perform the procedure described above. The computer program product thus possesses the advantages and features of the described procedure. A computer program product, such as a computer program tool, can be provided or delivered, for example, as a storage medium such as a memory card, USB flash drive, CD-ROM, DVD, or as a downloadable file from a server on a network. This can be achieved, for example, in a wireless communication network by transmitting a corresponding file containing the computer program product or the computer program tool.

[0046] According to one aspect of the invention, a driver assistance system for a vehicle is proposed. The driver assistance system is designed as at least one control unit. For example, the driver assistance system is a single control unit. For example, the driver assistance system is a circuit containing several control units. The driver assistance system is configured to execute the proposed method. The embodiments and features described for the proposed method apply accordingly to the proposed driver assistance system.

[0047] The driver assistance system can, for example, include an interface device for receiving at least one sensor signal. A sensor signal can, for example, be a signal from an environmental sensor, a position sensor, and / or an external vehicle sensor.

[0048] 10

[0049] The driver assistance system may, for example, include a processor for controlling a control flow and / or performing calculations. The driver assistance system may, for example, include a storage device for providing the trajectory data set and, if applicable, environmental data, such as a SLAM card. The driver assistance system may, for example, include an interface device for outputting the trajectory data set and / or a longitudinal and / or lateral control signal based on a comparison of the trajectory data set with at least one sensor signal.

[0050] Furthermore, a vehicle is proposed that incorporates a proposed driver assistance system. The embodiments and features described for the proposed method apply accordingly to the proposed vehicle.

[0051] The vehicle is, for example, a passenger car or a truck. The vehicle preferably includes a number of sensor units designed to detect the vehicle's driving status and its surroundings. Examples of such sensor units include imaging devices such as cameras, radar (radio detection and ranging), or lidar (light detection and ranging), ultrasonic sensors, positioning sensors, wheel angle sensors, and / or wheel speed sensors. Each sensor unit is configured to output a sensor signal, for example, to the driver assistance system, which performs semi-autonomous or fully autonomous driving based on the detected sensor signals.

[0052] A vehicle's level of automation, for example, corresponds to an automation level according to the SAE classification system. The SAE classification system was published in 2014 by SAE International, a standards organization for motor vehicles, as J3016, "Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems." It is based on six different levels of automation and considers the degree of system intervention required and the driver's attention required. The SAE automation levels range from level 0, which corresponds to a 2024PF02762

[0053] 11

[0054] The automation levels range from fully manual systems, through driver assistance systems at levels 1 and 2, to semi-autonomous (levels 3 and 4) and fully autonomous (level 5) systems, where no driver is required. An autonomous vehicle (also known as a driverless car, self-driving car, and robotic car) is a vehicle capable of perceiving its surroundings and navigating without human input, corresponding to SAE automation level 5.

[0055] Other possible implementations of the invention also include combinations of features or embodiments described previously or subsequently with regard to the exemplary embodiments, even if not explicitly mentioned. In such cases, the person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the invention.

[0056] Further advantageous embodiments and aspects of the invention are the subject of the dependent claims and the exemplary embodiments of the invention described below. The invention will be explained in more detail below with reference to preferred embodiments and the accompanying figures.

[0057] Fig. 1 schematically shows a top view of a vehicle with environmental sensors;

[0058] Fig. 2 shows a flowchart of a proposed method for operating a driver assistance system, including optional and / or mutually alternative steps for illustrative purposes; and

[0059] Fig. 3 shows four partial steps of processing a trajectory data set using the proposed method (including optional steps) based on a planar projection of waypoints.

[0060] In the figures, identical or functionally equivalent elements have been given the same reference symbols unless otherwise indicated. 2024PF02762

[0061] 12

[0062] Fig. 1 shows a schematic bird's-eye view of a vehicle 100. The vehicle 100 is, for example, a car located in an environment 140. The car 100 has a control unit 110, which is a driver assistance system or, in an operating state, acts as a driver assistance system. Furthermore, a plurality of environmental sensor devices 120, 130 are arranged on the car 100, including, for example, optical sensors 120 and ultrasonic sensors 130. The optical sensors 120 include, for example, visual cameras, radar, and / or lidar. The optical sensors 120 can each capture an image of a respective area from the environment 140 of the car 100 and output it as an optical sensor signal. The ultrasonic sensors 130 are configured to detect the distance to objects located in the environment 140 and to output a corresponding sensor signal.Using the sensor signals detected by sensors 120 and 130, the control unit 110 is able to drive the car 100 semi-autonomously or even fully autonomously. In addition to the sensors shown in the figure...

[0063] In addition to the optical sensors 120 and ultrasonic sensors 130 shown in Figure 1, the vehicle may also have 100 different additional sensor devices 120, 130. Examples include a microphone, an accelerometer, an antenna with a coupled receiver for receiving electromagnetically transmitted data signals, and the like.

[0064] Next, a process 200 for operating the driver assistance system 110 will be briefly described with reference to Fig. 2. Reference is made to the illustration in Fig. 3, which illustrates certain parts of the process 200 in four views.

[0065] Procedure 200 can be triggered by various events. For example, in step S201, further steps S205 to S220 of procedure 200 are triggered by the end of a recording of a trajectory data set 300. Similarly, in step S202, further steps S205 to S220 of procedure 200 are triggered by approaching at least one of several waypoints of the trajectory data set 300. This approach can be triggered, for example, by a comparison of the 2024PF02762

[0066] 13

[0067] Trajectory data set 300 is identified with an actual position. The approach can be detected, for example, by comparing a navigation system's route guidance function with trajectory data set 300. For instance, in step S203, the subsequent steps S205 to S220 of procedure 200 are triggered by the end of a follow along the trajectory. Thus, trajectory data set 300 was recorded in a previous recording run, which vehicle 100 followed in the last run. Additional and / or different environmental features may have been detected, so that adjusting the waypoints to the environment can improve future runs 140. For example, in step S204, the subsequent steps S205 to S220 of procedure 200 are triggered by the expiration of a time interval; this can also be described as a cyclical execution of procedure 200.

[0068] It may be optional to check for the existence of a further start condition for the following steps in a separate step and / or in one of the trigger steps S201-S204. For example, procedure 200 may only be executed if the vehicle 100 is stationary. For example, procedure 200 may only be executed if the driver assistance system 110 is sufficiently underutilized.

[0069] In step S205, the trajectory data set 300 is provided. The data set describes several waypoints 302. The waypoints 302 define a trajectory 304, which is shown as a line in parts AD of Fig. 3. Fig. 3A illustrates the trajectory 304, showing, for example, vehicle 100 at its beginning. The waypoints 302 have a regular interval 306. The regular interval 306 corresponds to the length of vehicle 100.

[0070] In step S206, the trajectory data set 300 is searched for a segment 308 between two waypoints 302b and 302c, where the distance 310 between these waypoints is shorter than the regular distance 306. The segment 308 found in this search is named 2024PF02762

[0071] 14

[0072] Hereinafter referred to as "first" section 308. Section 308 extends from the first waypoint 302b to the second waypoint 302c. There is preferably another waypoint between these waypoints 302b and 302c.

[0073] In sub-step S207, each waypoint 302 is virtually superimposed by a control contour 312. In a simple case, the control contour 312 can be a circle. The circle can, for example, have a diameter of a regular interval 306. In the illustrated case, the control contour 312 is assumed to be the outline of the vehicle 100. The outline is a projection onto a road surface. The outline can be simplified, for example, in the form of a rounded rectangle. A control contour 312 is always positioned in the same way as the waypoint 302 it superimposes; for example, the centroid of the control contour 312 is superimposed over the waypoint 302. For non-circular control contours 312, the control contour 312 may be aligned with the trajectory 304 described by the waypoints 302.For example, the outline-shaped control contour 312 is aligned such that a corresponding longitudinal axis at waypoint 302 is tangential to the trajectory 304 described by waypoint 302.

[0074] In step S208, it is detected whether two (or more) control contours 312 overlap. Configurations may exist where every overlap 314 is to be detected. This is advantageous if each control contour 312 is slightly smaller than the regular spacing 306. In a preferred case, also shown in Fig. 3, an overlap 314 is only detected if two control contours 312 overlap by a predetermined amount. The predetermined amount is, for example, an overlap 314 of at least 10% (or, for example, 5% or 20%) of the area of ​​a control contour 312.

[0075] Fig. 3 Part B shows a superposition of outline-shaped control contours 312 over the waypoints 302, as well as several detected intersections 314.2024PF02762

[0076] 15

[0077] In step S209, a new section 316 is created. The new section 316 includes the first waypoint 302b and the second waypoint 302c as endpoints. In the case shown in Fig. 3, a section 318 (also containing the first waypoint 302b) aligns with the first section 308, and a section 320 (also containing the second waypoint 302c) aligns with the first section 308 with slight lateral and angular deviations. Therefore, the waypoints 302 between the first waypoint 302b and the second waypoint 302c are simply removed, see parts B and C of Fig. 3.

[0078] It may also happen that new waypoints 302 are calculated between the found waypoints 302b and 302c, for example, to obtain a smooth trajectory 304. For example, a Bézier curve is calculated from the first waypoint 302b to the second waypoint 302c, and then new waypoints 302 are determined along this curve.

[0079] The new section 316 preferably transitions smoothly from the existing previous section 318 into the new section 316 and further into the existing following section 320.

[0080] The trajectory 204 shown in Fig. 3 could, for example, have resulted from a back-and-forth journey in three passes. It has two waypoints 302, namely the second waypoint 302c and a middle waypoint 302e, at which the direction of travel of the recorded journey changed. Because the direction of travel changes twice in the identified section 308, the direction of travel is the same at the first waypoint 302b and the second waypoint 302c. However, if the direction of travel should not be the same, then it is advantageous that the newly generated section 316 contains an odd number of changes in direction. Thus, for example, a trajectory for reversing into a parking space can maintain a desired course.

[0081] In step S210, it is checked whether the newly created section 316 violates a drivability criterion. For example, based on available information, it is checked whether the newly created section 316 runs on a roadway and / or an object in the vicinity. 2024PF02762

[0082] 16

[0083] 140 does not get too close. This check can be based, for example, on the positions of object features stored in the trajectory data set. An object feature can be understood as an object component detected in a sensor signal. The check can be based, for example, on an externally provided and / or generated environment map by a vehicle assistance system. For example, the map can be generated during the recording drive and / or the follow-up drive using SLAM technology.

[0084] A front waypoint 302a is located along trajectory 304, immediately before the first waypoint 302b, according to trajectory data set 300. A back waypoint 302d is located along trajectory 304, immediately after the second waypoint 302c, according to trajectory data set 300. In step S211, another new section 322 is created, extending from and including front waypoint 302a to and including the second waypoint 302c. In step S212, another new section 324 is created, extending from and including front waypoint 302a to and including back waypoint 302d. In step S213, another new section 326 is created, extending from and including the first waypoint 302b to and including back waypoint 302d. Part C of Fig. 3 illustrates the positions of these different sections, although for illustrative reasons only the new section 316 generated in S209 is shown in the course of trajectory 304 in Part C.

[0085] The drivability check in S210, for example, follows each of steps S209 and S211-S213. However, the drivability check can also be integrated into each of these steps S209, S211-S213. It is also possible that the drivability check is not necessary, for example, because the respective new section is generated based on sensor data.

[0086] In step S214, one of the new sections 316, 322, 324, or 326 is selected. The selection criterion is, for example, the smoothness of the respective trajectory 304. Smoothness can be defined, for instance, as the change along trajectory 304 of the distance of a waypoint 302 from an imaginary connecting line between 2024PF02762.

[0087] 17

[0088] whose two adjacent waypoints 302 measure. It can also be said that from the four new sections 316, 322, 324, 326, the section 326 should be selected which leads to the smallest steering movements while driving along the resulting trajectory 304.

[0089] In step S215, the first section 308 is replaced by the new section 316. Considering steps S211 to S214, it can be said that the waypoints 302 of trajectory 304, from the first waypoint (e.g., 320b) of the new section 316, 322, 324, 326 to the last waypoint (e.g., 302d) of the new section 316, 322, 324, 326, are replaced by the new section 316, 322, 324, 326.

[0090] In step S216, steps 211 to 214 are repeated for the sequence of waypoints 302 resulting from the selection in S214. Depending on the section 316, 322, 324, or 326 previously selected in S214, a waypoint located further ahead may be used as the starting waypoint in S211 and S212, and / or a waypoint 302 located further back may be used as the ending waypoint in S212 and S213. In other words, the first waypoint 302 of the section selected in S214 functionally replaces the first waypoint 302b found in S206, and the last waypoint 302 of the section selected in S214 replaces the second waypoint 302c found in S206. One can say that in S216 a dependency on waypoint 302 is changed in order to repeat steps S211 to S213 with a changed dependency on the waypoints.

[0091] For illustration, reference is again made to Fig. 3, Part D. In the illustrated example, the new section 326 was selected in S214. To determine whether a larger, newly generated section is better, waypoint 302a is used as the front waypoint and waypoint 302b as the first waypoint for the new iteration, but waypoint 302d is used as the second waypoint and a subsequent waypoint 302f as the back waypoint. Step S216 can be viewed as a recursive iteration. Alternatively, step S216 can be viewed as a redefinition of waypoints 302 according to the selection in S214. 2024PF02762

[0092] 18

[0093] The last selected section (section 326 in the case of Fig. 3, part D) can be included in the selection instead of step S209, in order to use the same selection algorithm in S214. It can be specified that the loop in S216 terminates after a certain number of repetitions. It can also be specified that the loop in S216 terminates if no other selection is made in S214.

[0094] In step S217, the first section may not be replaced by a new section. For example, no new section 316, 322, 324, 326 may be recognized as drivable in S210, so changing the trajectory data set would offer no advantage.

[0095] It is possible that in step S218 a further search is initiated to search the trajectory data record 300 in S206 for another "first" section 308 between two waypoints 302, for which the distance 310 is shorter than the regular distance 306. In this exemplary way, the entire trajectory 304 can be improved.

[0096] In step S219, the (modified) trajectory data set 300 is saved.

[0097] Vehicle 100 may follow the (modified) trajectory 304 at a later time or never. It is also possible that, in a subsequent step S220, vehicle 100 follows trajectory 304, which is defined by the waypoints 302 of the trajectory data set 300. In doing so, vehicle 100 or the driver assistance system 110 may detect further features in the environment 140 from received signals from the environmental sensors 120, 130. Therefore, it is advantageous to then trigger the remaining steps of procedure 300 again in S203. 2024PF02762

[0098] 19

[0099] In a preferred embodiment of method 200, a first section 308 is detected in S206 only if, according to the trajectory data set 300, there is at least one further waypoint 304e between waypoints 302b and 302c, and method 200 does not have steps S211 to S214. In a preferred embodiment of method 200, a first section 308 is detected in S206 regardless of whether, according to the trajectory data set 300, there is at least one further waypoint 304e between waypoints 302b and 302c, and method 200 has at least some of steps S211 to S214. In a preferred embodiment of method 200, a first section 308 is recognized in S206 only if, according to the trajectory data set 300, there is at least one further waypoint 304e between waypoints 302b, 302c, and method 200 has at least part of steps S211 to S214.

[0100] Although the present invention has been described using exemplary embodiments, it can be modified in many ways. 2024PF02762

[0101] 20

[0102] REFERENCE MARK LIST

[0103] 100 vehicles

[0104] 110 Control unit

[0105] 120 optical sensor

[0106] 130 Ultrasonic sensor

[0107] 140 surroundings

[0108] 200 methods for processing a trajectory data set

[0109] 300 trajectory data set

[0110] Waypoint 302

[0111] 302a front waypoint or waypoint before first waypoint

[0112] 302b first waypoint

[0113] 302c second waypoint

[0114] 302d back waypoint or waypoint after second waypoint

[0115] 302e middle waypoint

[0116] 302f rear waypoint

[0117] 304 Trajectory

[0118] 306 regular interval

[0119] 308 first section

[0120] 310 shorter distance

[0121] 312 Control contour

[0122] 314 Overlap

[0123] 316 new section from the first waypoint to the second waypoint

[0124] 318 Section before first section

[0125] 320 Section after first section

[0126] 322 new section from front waypoint to second waypoint

[0127] 324 new section from front waypoint to back waypoint

[0128] 326 new section from first waypoint to rear waypoint

[0129] S201 Triggering the next steps of the procedure by ending a recording of the trajectory data set 2024PF02762

[0130] 21

[0131] 5202 Triggering the next steps of the procedure by approaching at least one of the waypoints of the trajectory data set

[0132] 5203 Triggering the next steps of the procedure by the end of a follower along the trajectory

[0133] 5204 Triggering the further steps of the procedure at regular time intervals

[0134] 5205 Providing a trajectory dataset that has multiple waypoints and a regular spacing between waypoints

[0135] 5206 Searching the trajectory dataset for a segment between two waypoints whose distance is shorter than the regular distance,

[0136] 5207 Virtual overlay of each waypoint with a control contour

[0137] 5208 Detecting an overlap of two control contours

[0138] 5209 Creating a new segment from the first waypoint to the second waypoint

[0139] 5210 Check if the newly created section violates a drivability criterion. 5211 Create a new section from a waypoint directly before the first waypoint to the second waypoint.

[0140] 5212 Create a new segment from the waypoint immediately before the first waypoint to a waypoint immediately after the second waypoint

[0141] 5213 Create a new segment from the first waypoint to the waypoint immediately after the second waypoint

[0142] 5214 Selecting one of the new sections depending on at least one smoothness of the respective trajectory

[0143] 5215 Repeating the creation of new sections and selecting a new section with modified waypoints

[0144] 5216 Replacing the first section with the new section

[0145] 5217 Not replacing the first section with a new section

[0146] 5218 Repeat for another first section

[0147] 5219 Saving the trajectory data set

[0148] 5220 descendants along the waypoints of the trajectory dataset

Claims

2024PF02762 22 PATENT CLAIMS 1. Computer-implemented method (200) for operating a driver assistance system (110), comprising: Providing (S205) a trajectory dataset (300) containing several waypoints (302) of a trajectory (304), wherein the trajectory (304) has a regular interval (306) between successive waypoints (302) in at least part of the trajectory; searching (S206) the trajectory dataset (300) for a first segment (308) from a first waypoint (302b) to a second waypoint (302c) for which two waypoints (302b, 302c) the distance (310) from the first waypoint (302b) to the second waypoint (302c) is shorter than the regular interval (306); generating (S209) a new segment (316) for the trajectory (304) from the first waypoint (302b) to the second waypoint (302c), and Replacing (S215) the first section (308) with the new section (316).

2. Method according to claim 1, characterized in that the search (S206) includes: virtual overlay (S207) of each waypoint (302) by a control contour (312), and Recognizing (S208) an intersection (314) of two control contours (312) as the shorter distance (310) of the first waypoint (302b) to the second waypoint (302c).

3. Method according to claim 2, characterized in that each intersection (314) of two control contours (312) is recognized as the shorter distance (310) of the first waypoint (302b) to the second waypoint (302c).

4. Method according to claim 2, characterized in that only an overlap (314) of two control contours (312) by a predetermined area dimension is recognized as the shorter distance (310) of the first waypoint (302b) to the second waypoint (302c). 2024PF02762 23 5. Method according to one of claims 2 to 4, characterized in that the regular distance (306) corresponds to a length of the vehicle (100) which has recorded the waypoints (302) of the provided trajectory data set (300), and wherein preferably the control contour (312) corresponds to an outline of this vehicle (100).

6. Method according to one of the preceding claims, characterized in that the method (200) contains, before the replacement (S215): a) Generating (S211) another new segment (322) from a waypoint (302a) directly before the first waypoint (302b) to the second waypoint (302c), b) Generating (S212) another new segment (324) from the waypoint (302a) immediately before the first waypoint (302b) to a waypoint (302d) immediately after the second waypoint (302c), and / or c) Generating (S213) another new segment (326) from the first waypoint (320b) to the waypoint (302d) directly after the second waypoint (302c), and selecting (S214) one of the new segments (316, 322, 324, 326) depending on at least one smoothness of the respective trajectory (304).

7. Method according to claim 6, characterized in that the generation (S211 , S212, S213) of at least one further section (322, 324, 326) and the selection (S214) of one of the new sections (316, 322, 324, 326) are repeated.

8. Method according to one of the preceding claims, characterized in that the method (200) includes, before replacement (S215): checking (S210) whether the newly created section (316, 322, 324, 326) violates a drivability criterion.

9. Method according to one of the preceding claims, characterized in that when generating (S209, S211, S212, S213) a new section (316, 322, 324, 326), the direction of travel of the recording vehicle (100) is maintained at the first waypoint (302a, 302b) and at the last waypoint (302c, 302d) of the respective section (316, 322, 324, 326). 2024PF02762 24 10. Method according to one of the preceding claims, characterized in that at least some of the steps (S205-S220) are triggered (S204) at regular time intervals.

11. Method according to one of the preceding claims, characterized in that at least part of the steps (S205-S220) are triggered (S201) by an end of a recording of the trajectory data set (300).

12. Method according to one of the preceding claims, characterized in that at least part of the steps (S205-S220) are triggered (S202) by approaching at least one of the waypoints (302) of the trajectory data set (300).

13. Computer program product comprising instructions which, when the program is executed by a computer, cause it to execute the method (200) according to any one of claims 1-12.

14. Driver assistance system (110) for a vehicle (100) which is configured to perform the method (200) according to one of claims 1 - 12.

15. Vehicle (100) comprising a driver assistance system (110) according to claim 14.

Images