AUTONOMOUS PARKING METHOD

Abstract

Method (V1) for generating a parking trajectory (120) for autonomously parking a vehicle (100), comprising the steps: providing a path; detecting (S108) a defect (122) within an area (126) of the path; generating (S116) a parking trajectory (120) that can be followed by the vehicle (100) based on the path and taking into account the detected defect (122).

Description

[0001] The present invention relates to a method for generating a parking trajectory for autonomously parking a vehicle, a data set for autonomously parking a vehicle based on a parking trajectory, a method for autonomously parking a vehicle based on a parking trajectory, a computer program product, a control device and a vehicle.

[0002] DE 10 2019 131 630 A1 discloses a method for avoiding a collision during a replay of a trained journey, comprising the steps of: training a trajectory with the ego vehicle; and providing at least one option for selecting at least one area of ​​a vehicle environment in order to limit the trajectory for automatic driving, wherein the selection of the area determines that the area for an evasive trajectory for collision avoidance is not available during a replay of the trajectory; starting the replay, detecting an obstacle during the replay, determining an evasive trajectory, and if no drivable evasive trajectory can be determined, stopping in front of the obstacle.

[0003] Furthermore, US 2021 / 0 284 183 A1 discloses an electronic control device for automotive application, which is configured to store repeated low-speed maneuvers including location data, among other things, for obstacles and associated dimensions in the environment; and to calculate an optimal path taking into account, among other things, the stored obstacles and associated dimensions and new obstacles that have appeared subsequent to the storage of repeated low-speed maneuvers.

[0004] Against this background, one object of the present invention is to provide means for comfortable autonomous parking in the presence of a damaged area of ​​a drivable surface, such as a pothole.

[0005] Accordingly, a method for generating a parking trajectory for autonomous parking of a vehicle is proposed.

[0006] The method for generating a parking trajectory for autonomous parking of a vehicle includes the following steps: Providing a path, detecting a damage site within an area of ​​the path, and generating a parking trajectory that can be followed by the vehicle based on the path and taking into account the detected damage.

[0007] The path is, for example, a trained path previously traveled by the same vehicle or another vehicle. Alternatively or additionally, the path is a path transmitted by a server or another vehicle. Alternatively or additionally, the path is transmitted to the vehicle, at least in sections, by an AVP system (Automated Valet Parking System). Preferably, the damage site is only recognized as such when it reaches a certain size, in particular 5, 10, 20, or 30 cm. 2exceeds. The damaged area is, for example, road damage and / or a pothole. A detected damaged area can, for example, already be provided with the path. Alternatively or additionally, the damaged area can only be detected when the parking trajectory is followed and taken into account in real time. Within an area of ​​the trajectory means in and preferably next to a driving path, so that avoidance trajectories are also included. Generating the trajectory means, in particular, generating the overall traceable parking trajectory at least section by section. This can, for example, take place piece by piece in real time during the follow-up drive. The tracking of the parking trajectory takes place, for example, in and / or against a recording direction.

[0008] Optionally, when generating the retrievable parking trajectory, the damaged area can be taken into account in such a way that the damaged area is not driven over and / or avoided. This has the advantage of allowing for more comfortable retrieval of the parking trajectory.

[0009] Optionally, if the provided path does not contain a damaged area, the damaged area is taken into account when generating the reversible parking trajectory in such a way that the damaged area is not driven over and / or avoided.

[0010] This takes previously unknown damage areas into account.

[0011] Optionally, if the provided path contains a damaged area that would be crossed over when following the path with the vehicle, the damaged area is taken into account when generating the reversible parking trajectory in such a way that the damaged area would be crossed over when following the reversible trajectory with the vehicle.

[0012] Optionally, it is provided that the at least one damaged area that was provided in the provided path and would be crossed over when following the path with the vehicle is taken into account when generating the traceable parking trajectory in such a way that the damaged area would be crossed over when following the traceable trajectory with the vehicle, and in particular the at least one damaged area that was only detected during or shortly before generating the traceable parking trajectory is taken into account when generating the traceable parking trajectory in such a way that the damaged areas are not driven over and / or avoided.

[0013] Preferably, damaged areas that have already been driven over during training are also driven over when following up and newly identified damaged areas are avoided.

[0014] The proposed method optionally includes: continuously and / or repeatedly detecting a position of a vehicle; recording a path traveled by the vehicle based on the detected positions while parking the vehicle; detecting a drivable surface by means of at least one sensor device, in particular of the vehicle and / or an infrastructure; detecting the presence of a damaged area within a region of the drivable surface, which region is located in front of the vehicle in a direction of travel of the vehicle, based on at least one sensor signal from the at least one sensor device; recording a damaged area position of each detected damaged area of ​​the drivable surface; generating a parking trajectory that can be followed by the vehicle based on the recorded path;and generating a data set containing the parking trajectory and all recorded damage location positions, and configured to contain the corresponding damage location for each of several damage locations. The proposed method therefore provides for recording a damage location position for each detected damage location along with a parking trajectory. The generated data set thus contains information about how the driver behaves when encountering a specific pothole. The method is thus suitable for defining a level of driving comfort expected by the driver or acceptable to the driver. The infrastructure is preferably a parking infrastructure, in particular an AVP (automated valet parking) infrastructure.

[0015] The area in a direction of travel in front of the vehicle is an area that is first detected by at least one of the vehicle's sensor devices and then driven through if the vehicle does not stop. If the vehicle's direction of travel is changed, for example, by changing the steering angle or switching between forward and reverse movement, the area in front of the vehicle in the changed direction of travel also changes.

[0016] The generated parking trajectory is preferably a trained parking trajectory. For example, a control device of a vehicle is configured to record and store a manually driven parking trajectory in a training mode. For example, various sensor signals are recorded that characterize a driving state of the vehicle, such as speed, position, steering angle, and the like, as clearly as possible. In addition, sensor signals from the vehicle's environmental sensors are recorded, which, for example, enable an image of the vehicle's surroundings, in particular the position of obstacles in the surroundings. By playing back the vehicle's driving state synchronously, i.e., repeating it, the trained parking trajectory can be retraced.

[0017] Optionally, the detection of the damaged area can include: recording at least one extent, in particular an extent in a plane of the drivable surface and / or a depth, of each detected damaged area; wherein the generated data set contains all recorded extents of the damaged areas and is configured to contain the associated at least one extent for each of the multiple damaged areas. Thus, an extent of the damaged area that influences driving comfort can be documented. This method option therefore has the benefit of being able to quantitatively assess or predict driving comfort when following the parking trajectory. For the purpose of a computationally simple definition, it can be provided to determine the extent in a plane of the drivable surface along the parking trajectory and / or transversely to the parking trajectory.For the purpose of an accurate determination, it may be intended to determine a furthest extent in any direction.

[0018] Optionally, the method may include: providing a classification scheme for classifying damaged areas of the drivable surface based on at least one characteristic value of the damaged area, which characteristic value can be detected by the at least one sensor device; and classifying each detected damaged area into a damaged area class by applying the classification scheme to the at least one sensor signal; wherein the generated data set for each classified damaged area contains the damaged area class and is configured to contain the associated damaged area class for each of the multiple damaged areas. Classification creates easy comparability. This is advantageous because classification generally disregards minor changes to a damaged area, in particular a pothole, which are to be expected on a regular basis.The classification scheme is preferably a set of rules that can be applied to a signal from the at least one sensor device. The drivable surface is preferably a road surface and / or a grass surface and / or a gravel surface.

[0019] According to a next aspect of the invention, a data set for autonomously parking a vehicle based on a parking trajectory is proposed. The proposed data set contains a parking trajectory that can be followed by the vehicle. The proposed data set is configured to contain, in addition to the parking trajectory, an associated damage location for each of a plurality of damage locations, as well as preferably at least one associated extent and / or an associated damage location class. Each damage location indicates a damage location of a region of the drivable surface, which region is located in front of the vehicle during at least part of the journey when a vehicle is traveling along the parking trajectory in a direction of travel of the vehicle.In addition to the parking trajectory, the dataset contains information about whether it was acceptable for a driver to drive over a specific damage area during a recorded drive, as well as, if applicable, what type of damage area was acceptable for the driver. The dataset is therefore suitable for defining the level of driving comfort expected by the driver or acceptable for the driver.

[0020] According to a next aspect of the invention, a method for autonomously parking a vehicle based on a parking trajectory is proposed. The proposed method includes providing a data set containing a parking trajectory that can be followed by a vehicle, wherein the data set is configured to contain a plurality of damage point positions, wherein each damage point position indicates a damage point of a region of a drivable surface, which region is located in front of the vehicle in a direction of travel of the vehicle at least during part of the journey when a vehicle is traveling along the parking trajectory. The proposed method includes detecting a region of a drivable surface located in front of the vehicle in the direction of travel of the vehicle along the parking trajectory by means of at least one sensor device of the vehicle.The proposed method includes checking the drivability of the area in front of the vehicle, comprising the steps of: detecting the presence of a damaged area within the area of ​​the drivable surface in front of the vehicle based on a sensor signal from the sensor device; and determining that the area in front of the vehicle is not drivable if a distance between a damaged area position of a detected damaged area and a damaged area position contained in the data set is greater than a preset position threshold. The proposed method includes autonomous driving or following the vehicle along the parking trajectory if the area in front of the vehicle is drivable; and initiating an autonomous avoidance of the detected damaged area otherwise.The method proposed here allows the driver to continue driving over a pothole or other damaged area on the drivable surface if the pothole / damaged area was already present during the recording. This prevents the vehicle from unexpectedly deviating from the trained parking trajectory, thus achieving a high level of driving comfort. However, if a new pothole / other damaged area is detected by the sensor device, the vehicle initiates a detour around the pothole / other damaged area. This prevents the driver from being exposed to the vehicle's reaction to an unexpected pothole, and a high level of driving comfort can be maintained.

[0021] In a further development, it can be provided that the provided data set is configured to contain the associated extent for each of the multiple damage areas; and that checking the drivability includes: detecting an extent of the detected damage area based on the signal from the at least one sensor device; and determining that the area in front of the vehicle is not drivable if the detected extent is greater than the extent contained in the data set and associated with the damage area by more than a preset extent threshold. The method is thus capable of preventing driving over a damage area / pothole that has become larger than a recording point and is therefore potentially perceived by a driver as a comfort impairment.

[0022] In a further development, it can be provided that the provided data set is configured to contain an associated damage class for each of the multiple damage locations; and that the method includes: providing a classification scheme for classifying a damage location based on at least one characteristic value of the damage location, which characteristic value can be detected by the at least one sensor device; wherein checking the drivability includes: classifying the detected damage location into a damage class by applying the classification scheme to the at least one sensor signal; and determining that the area in front of the vehicle is not drivable if the damage class of the detected damage location is not the same as a damage class stored in the data set for the detected damage location. Simple comparability is created by means of the classification.This is advantageous because classification generally disregards minor changes in a damaged area, especially a pothole, which are to be expected on a regular basis. The classification scheme is preferably a set of rules that can be applied to a signal from the at least one sensor device.

[0023] Optionally, the method may include: detecting the presence of another damaged area within a region of the drivable surface adjacent to the region of the drivable surface located in front of the vehicle when the vehicle is traveling along the parking trajectory, based on a sensor signal from the sensor device; wherein the autonomous avoidance of the additional damaged area is also initiated. Initiating the avoidance thus simultaneously affects the detected damaged area in the region located in front of the vehicle when traveling along the parking trajectory and the detected damaged area in the adjacent area. In other words, a parking trajectory is sought to simultaneously avoid all detected, inaccessible damaged areas and all detected obstacles.

[0024] Optionally, the method may include detecting the absence of a damage at a damage location contained in the data set; and updating the data set based on the detected absence. Thus, a potential repair can be taken into account, resulting in a more recent road damage at the same or a similar location, which was not tolerated by the driver in the original recording, not being driven over.

[0025] It may further be the case that checking the drivability includes: detecting the presence of an obstacle in, on, and / or above the area located in front of the vehicle along the parking trajectory and / or a bypass trajectory based on a sensor signal from the at least one sensor device; and determining that the area in front of the vehicle is not drivable if at least one obstacle is present in the area in front of the vehicle; wherein the autonomous bypassing of the obstacle is initiated. The method is therefore suitable for initiating bypassing a new or grown damaged area and bypassing another obstacle using very similar means. The method can thus be implemented efficiently and executed quickly, leaving computing power for additional convenience functions.

[0026] Furthermore, a computer program product is proposed, which comprises instructions that, when executed by a computer, cause the computer to execute the above-described method for generating a parking trajectory and / or the above-described method for autonomously parking a vehicle. The embodiments and features described for the proposed method apply accordingly to the proposed computer program product.

[0027] A computer program product, such as a computer program means, can be provided or delivered, for example, as a storage medium, such as a memory card, USB stick, CD-ROM, DVD, or in the form of a downloadable file from a server in a network. This can be done, for example, in a wireless communications network by transmitting a corresponding file with the computer program product or the computer program means.

[0028] According to another aspect of the invention, a control device for a vehicle is proposed. The proposed control device is configured to execute the above-described method for generating a parking trajectory and / or the above-described method for autonomously parking a vehicle. The embodiments and features described for the proposed method apply accordingly to the proposed control device.

[0029] The control device is configured, in particular, for fully autonomous driving of the vehicle. Fully autonomous driving is understood, for example, to mean that the parking assistance system controls a steering device and / or an automatic gearshift, as well as a drive device and a braking device.

[0030] According to another aspect of the invention, a vehicle is proposed that includes at least one sensor device arranged and configured to detect a surface traversable by the vehicle; at least one actuator configured to specify a direction of travel and / or speed of the vehicle; and the above-proposed control device, which is connected in a communicable manner to the at least one sensor device and the at least one actuator. The embodiments and features described for the proposed method apply accordingly to the proposed vehicle.

[0031] Further possible implementations of the invention also include combinations of features or embodiments described above or below with respect to the exemplary embodiments that are not explicitly mentioned. In this case, the person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the invention.

[0032] 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 using preferred embodiments with reference to the accompanying figures. Fig. 1 shows schematically in plan view a vehicle according to an embodiment of the invention; Fig. 2 shows a schematic flowchart of a method for generating a parking trajectory for autonomous parking of a vehicle according to the embodiment of the invention; Fig. 3 schematically shows in a flowchart a method for autonomously parking a vehicle based on a parking trajectory according to the embodiment of the invention; Fig. 4 shows schematically in plan view a first scenario in which the methods according to the embodiment of the invention are applied; Fig. 5 shows schematically in plan view a second scenario in which the methods according to the embodiment of the invention are applied; Fig. Figure 6 shows schematically in plan view a third scenario in which the methods according to the embodiment of the invention are applied; and Fig. Figure 7 shows schematically in plan view a fourth scenario in which the methods according to the embodiment of the invention are applied.

[0033] In the figures, identical or functionally equivalent elements have been given the same reference numerals unless otherwise stated.

[0034] The Fig. 1 shows a schematic view of a vehicle 100 from a bird's eye view. The vehicle 100 is, for example, a car located in an environment 200. The vehicle 100 has a parking assistance system 102, which is embodied, for example, as a control device. Furthermore, a plurality of environmental sensor devices 104, 106 are arranged on the vehicle 100, which are, for example, optical sensors 104 and ultrasonic sensors 106. The optical sensors 104 include, for example, visual cameras, a radar, and / or a lidar. The optical sensors 104 can each capture an image of a respective area from the environment 108 of the vehicle 100 and output it as an optical sensor signal. The ultrasonic sensors 106 are configured to detect a distance to objects located in the environment 108 and to output a corresponding sensor signal.Using the sensor signals detected by sensors 104, 108, parking assistance system 102 is capable of driving vehicle 100 semi-autonomously or fully autonomously. In addition to the functions described in the . Fig. In addition to the optical sensors 104 and ultrasonic sensors 106 shown in Figure 1, the vehicle 100 may be provided with various additional sensor devices 104, 106. Examples include a microphone, an acceleration sensor, an antenna with a coupled receiver for receiving electromagnetically transmittable data signals, and the like.

[0035] For example, a vehicle's level of automation has an automation level of 3 to 5 according to the SAE classification system. The SAE classification system was published in 2014 by SAE International, an automotive standards organization, 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 takes into account the degree of required system intervention and driver attention. The SAE automation levels range from level 0, which corresponds to a fully manual system, through driver assistance systems at levels 1 to 2, to partially autonomous (levels 3 and 4) and fully autonomous (level 5) systems, where a driver is no longer required.An autonomous vehicle (also known as a driverless car, self-driving car, and robotic car) is a vehicle capable of sensing its environment and navigating without human input and meets SAE Automation Level 5.

[0036] In the following, a method V1 for generating a parking trajectory 120 for autonomous parking of the vehicle 100 is described using the flow chart of Fig. 2 briefly described. Then, based on the Fig. 3 describes a method V2 for parking a vehicle 100 based on a parking trajectory 120. The effect of the two methods V1, V2 is illustrated by four Fig. 4 to 7 scenarios are described individually.

[0037] The scenarios of Fig. 4 to 7 have in common that the environment 108 is a property 110 or at least a part thereof. The property has a driveway 112. The driver and / or the individual vehicle 100 is assigned a specific parking space 116 from several parking spaces 114. The driveway 112 is drivably connected to the parking space 116 by a roadway 118. In all four scenarios shown, the driver initially controls a parking maneuver of the vehicle 100, during which the method V1 for generating a parking trajectory 120 for autonomously parking a vehicle 100 is executed. Furthermore, in all four scenarios, the driver later has a parking assistance system park the vehicle 100 autonomously. This parking assistance system is formed, for example, by the control device 102. Meanwhile, the control device 102 executes the method V2 for autonomously parking the vehicle 100 based on the parking trajectory 120.

[0038] In a step S100, a classification scheme for classifying damaged areas 122 is provided. The classification scheme is adapted to a characteristic value that can be detected by at least one of the sensor devices 104, 106 of the vehicle 100. Potholes 122, in particular, are considered as damaged areas. The classification scheme defines at least two classes of damaged areas 122, namely a class of passable damaged areas 122 and a class of damaged areas 122 that cannot be passed over without compromising comfort.

[0039] Comfort when driving through a damaged area 122 using the vehicle 100 can depend on several factors. Therefore, the provided classification scheme is preferably prepared to assess comfort when driving through at a low speed of no more than 50 km / h, preferably no more than 30 km / h, and more preferably no more than 20 km / h. For this reason, the provided classification scheme is preferably tailored to a chassis property of the vehicle, such as a spring travel, a spring constant, a damper constant, an axle weight, a natural frequency, and / or tires of the vehicle. The provision in step S100 can, for example, include selecting a classification scheme from a list of multiple classification schemes based on the chassis property and / or the vehicle 100.

[0040] In a step S102, a position of the vehicle 100 is repeatedly and / or continuously detected. The position can be detected, for example, by means of satellite navigation, by means of WLAN detection, by means of odometry, and / or by means of two- and / or three-dimensional detection of the environment 108.

[0041] In a step S104, a path traveled by the vehicle 100 during driving and parking of the vehicle 100 is recorded, based on the positions or position information detected in step S102.

[0042] In a step S106, while the vehicle 100 is being driven and parked, a drivable surface 124 is detected by means of at least one of the sensor devices 104, 106. In particular, a drivable surface 124 is detected in an area 126 that is located in front of the vehicle 100 in the direction of travel of the vehicle 100. The term "area in the direction of travel in front of the vehicle" preferably takes into account a forward and reverse direction and / or a steering angle. Furthermore, the "area in the direction of travel in front of the vehicle" can preferably be understood as a minimum requirement for a detection area, so that, for safety reasons, a drivable surface of at least one adjacent area 132 is additionally detected by the at least one sensor device 104, 106. According to one understanding, a digital representation of the detected drivable surface 124 can be generated in step S106.

[0043] In a step S108, for example, the control device 102 detects, on the basis of the drivable surface 124 detected in step S106, in particular on the basis of the digital representation of the detected drivable surface 124 generated in step S106, whether a damaged area 122 is present within the area 126 which is located in front of the vehicle 100 when the vehicle 100 travels along the path.

[0044] In a step S110, a damaged spot position of each damaged spot 122 of the drivable surface 124 previously detected in step S108 is recorded.

[0045] In step S112, the extent of the damaged area 122 detected in step S108 in a plane of the drivable surface 124 and the depth of the same damaged area 122, or its extent in a depth direction or vertical direction, are detected and recorded. The sensor signal detected in step S106 forms the basis for this.

[0046] In a step S114, the damaged area 122 detected in step S108 is classified. For this purpose, the classification scheme provided in step S100 is applied, for example, to the sensor signal acquired in step S106, for example, in the manner of a filter.

[0047] Steps S112 and S114 can be provided as alternatives to save computing time. Steps S112 and S114 can be used cumulatively to enable the damage site 122 to be described particularly precisely.

[0048] In step S116, a parking trajectory 120 is generated. For this purpose, a training method is applied to the path recorded in step S104. The generated parking trajectory 120 is preferably configured to be followed by the vehicle 100 and, more preferably, by any vehicle.

[0049] In step S118, a data set is generated. The data set is generated, for example, by creating a data structure containing the parking trajectory 120 as a sequence of positions and the damage areas 122 ordered by damage area position, to which the extent(s) and / or the damage area classification are preferably linked.

[0050] In a step S120, the generated data set is provided. For example, the data set may be stored in the control device 102 in a retrievable manner. For example, the data set may be stored on a server in a retrievable manner via a network.

[0051] Thus, executing method V1 during a parking maneuver of vehicle 100 results in a data set suitable for detecting, during a re-entry, the damaged areas, in particular potholes 122, which the driver accepted during the recorded drive. Therefore, the parking maneuver can be re-enacted with the same driving comfort as the parking maneuver.

[0052] The generated data set thus contains a traceable parking trajectory 120, and it is configured to contain an associated damage position, preferably at least one associated extent, and preferably a damage class for no damage, for one damage 122, and in particular for several damages 122.

[0053] Next, the Fig. 3 describes a method V2 for autonomously parking a vehicle 100 based on the parking trajectory 120. The method V2 includes, for example, steps S200 to S236.

[0054] In step S200, a data set created, for example, in method V1 is provided. For example, the data set is retrieved from the control device 102 or from a server via a network.

[0055] In step S202, a classification scheme is provided, as retrieved. The classification scheme is preferably identical to the classification scheme provided in step S100. For example, the classification scheme may be contained or identified in the data set.

[0056] In a step S204, a drivable surface 124 is detected by means of at least one sensor device, such as one of the sensor devices 104, 106. The at least one sensor device 104, 106 generates or generates at least one sensor signal, which is evaluated, for example, by the control device 102, whereby the control device 102 preferably detects a digital representation of the drivable surface 124. In step S204, at least one region 126 of the drivable surface 124 located in front of the vehicle 100 in the direction of travel of the vehicle 100 along the parking trajectory 120 is detected.

[0057] In a step S206, the navigability of the area 126 located in front of the vehicle 100 in the direction of travel is checked. The direction of travel preferably follows from the provided parking trajectory 120 and can take into account a forward direction, a reverse direction, and / or a steering angle. In this case, navigability preferably does not mean technical feasibility, but rather it preferably means a comfort criterion. For example, navigability can be defined as a driving permit to obtain a driving comfort that can be assessed based on the data set. Step S206 is preferably a collection of several substeps S208 to S226.

[0058] In a step S208, the presence of a damaged area 122 within the region 126 of the drivable surface 124 located in front of the vehicle 100 is detected. The detection is based on the at least one sensor signal from the at least one environmental sensor 104, 106. The detection in step S208 is preferably similar to the detection in step S108 and / or a detection underlying the data set provided in step S200. For example, the data set provided in step S200 can contain sensor parameters that were documented during a method for generating the data set, such as preferably during an execution of method V1.

[0059] If no presence of a damaged area 122 is detected in step S208, i.e. if no damaged area 122 is detected in the area in front of the vehicle 100, steps S210 to S218 are preferably skipped and / or the method continues with one of the steps S220 and / or S226

[0060] In step S210, for example, a damaged area position of the damaged area 122 detected in step S208 is determined and compared with the provided data set. For example, in this way, it is determined in this step S210 whether the area 126 in front of the vehicle 100 can be driven through with almost the same level of comfort as during a parking maneuver underlying the parking trajectory 120. Specifically, in step S210, the damaged area position of the detected damaged area 122 is compared with the data set based on a provided distance threshold. If, for example, the damaged area position of the detected damaged area 122 is not within a circle, whose radius is the provided distance threshold, around a damaged area position contained in the data set, the area 126 in front of the vehicle 100 is assessed as not being driveable.It is therefore assumed that the damaged area identified in step S208 was not present at the time of training the parking trajectory 120, so it is unknown whether this damaged area 122 seemed acceptable for the comfort of a driver of the vehicle 100. Therefore, it is decided that the vehicle 100 should avoid the damaged area.

[0061] Steps S212 to S226 described below are preferably only carried out if it is not determined in step S210 that the area in front of the vehicle 100 is not passable.

[0062] In step S212, an extent of the damaged area 122 identified in step S208 is detected. For this purpose, the same method as in step S112 is preferably used. The method can, for example, be documented in the data set.

[0063] In a step S214, the drivability of the area 126 located in front of the vehicle 100 is compared based on the extent detected in step S212 and an extent of the damaged area 122 contained in the provided data set at the time of the trained parking maneuver. If the detected current extent does not exceed the extent stored in the data set plus an extent threshold, the area 126 is considered drivable; otherwise, the area 126 located in front of the vehicle 100 is determined to be impassable.

[0064] Steps S216 to S226 described below are preferably only carried out if it is not determined in step S214 that the area 126 located in front of the vehicle 100 is not passable.

[0065] In a step S216, the damaged area 122 identified in step S208 is classified based on the classification scheme provided in step S202. The classification, as well as the classification scheme, preferably resembles the classification in step S114 and the classification scheme provided in step S100.

[0066] In step S218, the drivability of the area 126 located in front of the vehicle 100 is assessed based on the damage class assessed in step S216 and a damage class contained in the data set for the damage 122. If the damage classes do not match, it can be assumed that the damage 122 has deteriorated significantly and that driving comfort when driving over the damage 122 is reduced. Therefore, in this case, it is determined that the area 126 located in front of the vehicle 100 is not drivable with the comfort level of the training drive.

[0067] Steps S220 to S226 described below are preferably only carried out if it is not determined in step S218 that the area 126 in front of the vehicle 100 is not passable.

[0068] In a step S220, the absence of a damaged area 12 in the area 126 located in front of the vehicle 100 is detected, which damaged area 122 is recorded in the data set. This can be the case, in particular, after repair work. In this case, the provided data set is updated in a step S222. Preferably, an indication of the damaged area position is removed from the data set. This ensures that if a damaged area 122 is re-present at this damaged area position, it is recognized that this apparently re-detected damaged area 122 has newly occurred and was therefore not tolerated by the driver during the parking maneuver.

[0069] In a step S224, the presence of an obstacle 130 in and / or on and / or above the area 126 located in front of the vehicle 100 is detected. An obstacle 130 can be a collective term for any type of object that impedes travel. For example, the presence of a garbage can or the growth of a plant in a travel path can be detected.

[0070] In a step S226, it is determined that the area 126 in front of the vehicle 100 is not passable if an obstacle 130 is detected in, on and / or above this area 126 in step S224.

[0071] If it is not determined that the area 126 in front of the vehicle 100 is not passable, a step S228 is executed for autonomous driving of the vehicle 100. However, if it is determined, for example in one of the steps S210, S214, and / or S218, that the area 126 in front of the vehicle 100 is not passable, a step S232 is executed instead, which is preferably preceded by a step S230.

[0072] If the result of step S206 of checking the drivability of the area 126 located in front of the vehicle 100 is that the area 126 located in front of the vehicle 100 is drivable with the comfort of parking, the method V2 continues with step S228. In step S228, the vehicle 100 drives autonomously along the provided parking trajectory 120. For example, the vehicle 100 has at least one controllable actuator, which is arranged and prepared to specify a movement of the vehicle 100 and which is communicatively connected to the control device 102. The control device 102 executes a method for following the parking trajectory 120 and thus controls the at least one actuator. To follow the provided parking trajectory 120, it is desirable to take current environmental sensor data into account. Therefore, the control device 102 receives a sensor signal indicative of the environment.Control device 102 can receive this signal, for example, directly from one or more of the environmental sensors 104, 106 of the vehicle 100 and combine several sensor signals from different environmental sensors, or control device 102 can receive the sensor signal already in a pre-processed state, for example in the form of a digital environmental map, which preferably records currently detected obstacles and / or damaged areas.

[0073] A step S230 is preferably executed if it is determined in one of steps S210, S214, S218, and / or S226 that the area 126 located in front of the vehicle 100 is not passable. In step S230, the presence of another damaged area 134 in an adjacent area 132 is detected. The adjacent area 132 is adjacent to the area 126 located in front of the vehicle 100. Preferably, the adjacent area 132 even borders the area 126 located in front of the vehicle 100. In step S230, for example, at least one sensor signal from at least one of the environmental sensors 104, 106 is evaluated.

[0074] In a step S232, an autonomous bypass of the damaged area(s) 122 identified in steps S208 to S218 or in step S206 and in step S230 is initiated because it is determined in at least one of steps S210, S214, and S218 that the area 126 located in front of the vehicle 100 along the parking trajectory 120 cannot be accessed without compromising comfort. Step S232 preferably includes executing a method for determining a bypass trajectory.

[0075] In step S234, the vehicle 100 is stopped. Step S234 is executed, for example, if no detour trajectory 120 can be determined that meets the driving comfort requirements.

[0076] In a step S236, the detected damage location 122, 134 or the detected damage locations 122, 134 are autonomously avoided by the vehicle 100. The vehicle 100 is preferably controlled by the control device 102.

[0077] Above, an embodiment is described which contains several optional steps for both the training method V1 and the tracking method V2. Next, an application of the methods V1, V2 is explained using scenarios which are described in the Fig. 4 to 7 are each shown in a top view. Methods V1 and V2 are also generally suitable for application in different scenarios.

[0078] In the scenario of Fig. 4, a pothole 122 already exists in a drivable surface 124 of the roadway 118 at the time the driver executes the training method V1. The pothole 122 is an example of a damaged area of ​​the drivable surface 124. The pothole 122 is located in an area 126 that is located in front of the vehicle 100 in a direction of travel of the vehicle 100. Therefore, the pothole 122 is detected together with the drivable surface 124 in step S106 and recognized as a damaged area in step S108. The damaged area position of the pothole 122 is recorded in step S110, its extent in a plane of the drivable surface 124 and its depth are recorded in step S112, and it is classified in step S114. Thus, the data set generated in step S118 contains the information describing the pothole 122.

[0079] In the scenario of Fig. 4, it is further the case that the pothole 122 has not changed significantly at a time when the driver parks the vehicle 100 using the parking method V2. Therefore, in step S210, the damaged area position detected in step S208 and the damaged area position stored in the data set correspond very well, in step S214, the extent and depth determined in step S212 correspond to the extent and depth contained in the data set, and in step S218, the class determined in step S216 is equal to the class stored in the data set. Thus, in step S206, it is not determined that the area 126 located in front of the vehicle 100 along the parking trajectory 120 cannot be driven through without a loss of comfort. The control device 102 therefore steers the vehicle 100 closely along the parking trajectory 120 in step S228.

[0080] In the scenario of Fig. 5, at the time of method V1, there is no pothole 122 in the drivable area 124. However, at the time of method V2, the pothole 122 is located in the area 126 that is located in front of the vehicle 100 when the vehicle 100 is traveling along the parking trajectory 120. Therefore, in step S210, it is determined that the area 126 cannot be driven through with the same driving comfort as at the time of method V1. Therefore, in step S232, an autonomous avoidance of the detected damaged area 122 is initiated, and in step S236, the damaged area 122 is autonomously avoided along an avoidance trajectory 128.

[0081] In the scenario of Fig. 6, at the time of method V2, an obstacle 130 is present in the area 126 along the parking trajectory 120, the obstacle is detected in step S224, the non-driveability of the area 126 is determined in step S226, and method V2 proceeds to step S230. A neighboring area 132, which is adjacent to the area 126 along the parking trajectory 120, contains another pothole 134, which is a damaged area. Upon initiation of the autonomous avoidance in step S232 in method V2, the vehicle 100 autonomously avoids the obstacle 130 and the other pothole 134 along an avoidance trajectory 128 in step S236.

[0082] In the scenario of Fig.7, the pothole 122 exists at the time of method V1, and therefore, among other things, a depth is detected and recorded in step S112, and the pothole 112 is stored in the data set in step 118. Furthermore, at a time of method V2, the pothole 122 is at least partially filled with water due to rain. Therefore, the depth of the pothole 122 cannot be detected in step S212. Therefore, it cannot be determined in step S214 that the detected depth deviates from the depth stored in the data set by at most a depth threshold value. The result of step S206 is therefore that the area 126 along the parking trajectory 120 cannot be driven through without a loss of comfort. Therefore, an autonomous bypass of the damaged area 122 is initiated in step S232 and executed in S236.

[0083] For safety's sake, it should be noted at this point that the term "area 126 which is located in front of the vehicle 100 in a direction of travel of the vehicle 100 along the parking trajectory 120" refers to the time at which the method V1 or V2 is carried out.

[0084] The methods V1 and V2 and the data set thus provide means to implement parking along a parking trajectory with high driving comfort.

[0085] Although the present invention has been described using exemplary embodiments, it can be modified in many ways. LIST OF REFERENCE SYMBOLS 100 vehicles 102 Control device 104 optical sensor 106 Ultrasonic sensor 108 Surroundings 110 plot 112 Entrance 114 Parking 116 assigned parking spaces 118 Roadway 120 Parking trajectory 122 Pothole or damaged area 124 Road surface 126 Area located in front of a vehicle in a direction of travel 128 Avoidance trajectory 130 Obstacle 132 Area adjacent to an area along a parking trajectory 134 more potholes V1 Method for generating a parking trajectory for autonomous parking of a vehicle S100 Providing a classification scheme for classifying a damaged area S102 Detecting a position of a vehicle S104 Recording a driven path of a parking trip S106 Detection of an area of ​​a drivable surface in front of a vehicle S108 Detecting the presence of a damaged area in an area in front of a vehicle S110 Recording a detected damage location S112 Recording the extent of a detected damage site S114 Classifying a detected damage site based on a provided classification scheme S116 Generating a parking trajectory S118 Creating a data record S120 Providing a data set V2 Method for autonomous parking of a vehicle based on a parking trajectory S200 Providing a data set S202 Providing a classification scheme for classifying a damaged area S204 Detection of an area of ​​a drivable surface in front of a vehicle S206 Checking the passability of an area of ​​a passable surface in front of the vehicle S208 Detecting the presence of a damaged area in an area in front of a vehicle S210 Determining the drivability of an area in front of the vehicle based on a damage location of a detected damage location and a provided data set S212 Detecting the extent of a detected damage site S214 Determining the drivability of an area in front of the vehicle based on the extent of a detected damage site and a provided data set S216 Classifying a detected damage site based on a provided classification scheme S218 Determining the drivability of an area in front of the vehicle based on a classification of a detected damage site and a provided data set S220 Detecting the absence of a damaged area based on a provided data set and a sensor signal S222 Updating a provided record S224 Detecting the presence of an obstacle in an area in front of a vehicle S226 Determining the passability of an area in front of the vehicle based on a detected obstacle S228 autonomous driving along a provided parking trajectory S230 Detection of the presence of another damage site in an area adjacent to an area in front of the vehicle S232 Initiating an autonomous avoidance of an area in front of a vehicle S234 Stopping a vehicle S236 autonomous driving around an area in front of a vehicle QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] DE 10 2019 131 630 A1

[0002] US 2021 / 0 284 183 A1

[0003] Cited non-patent literature

[0000] J3016, “Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems

[0035]

Claims

[1] Method (V1) for generating a parking trajectory (120) for autonomous parking of a vehicle (100), comprising the steps: Providing a path; Detecting (S108) a damaged area (122) within a region (126) of the path; Generating (S116) a parking trajectory (120) that can be followed by the vehicle (100) on the basis of the path and taking into account the detected damage point (122). [2] Method according to claim 1, characterized by that when generating the reversible parking trajectory (120) the damaged area (122) is taken into account in such a way that the damaged area (122) is not driven over and / or avoided. [3] Method according to one of the preceding claims, characterized bythat if the provided path does not contain a damaged area (122), the damaged area (122) is taken into account when generating the reversible parking trajectory (120) in such a way that the damaged area (122) is not driven over and / or avoided. [4] Method according to one of the preceding claims, characterized by that if the provided path contains a damaged area (122) which would be crossed over when following the path with the vehicle (100), the damaged area (122) is taken into account when generating the followable parking trajectory (120) in such a way that the damaged area (122) would be crossed over when following the followable trajectory (120) with the vehicle (100). [5] Method according to one of the preceding claims, characterized bythat the at least one damaged area (122) which was provided in the provided path and which would be crossed over when following the path with the vehicle (100) is taken into account when generating the followable parking trajectory (120) in such a way that the damaged area (122) would be crossed over when following the followable trajectory (120) with the vehicle (100), and in particular the at least one damaged area (122) which was only detected when generating the followable parking trajectory (120) is taken into account when generating the followable parking trajectory (120) in such a way that the damaged areas (122) are not driven over and / or avoided. [6] Method according to one of the preceding claims, characterized by : Recording (S104) the path traveled by the vehicle (100) based on detected positions during parking of the vehicle (100); Detecting (S106) a drivable surface (124) by means of at least one sensor device (104, 106); Detecting (S108) the presence of the damaged area (122) within an area (126) of the drivable surface (124); Recording (S110) a damaged spot position of the detected damaged spot (112) of the drivable surface (124); Generating (S116) the parking trajectory (120) that can be followed by the vehicle (100) based on the recorded path; and Generating (S118) a data set containing the parking trajectory (120) and at least one damaged area position. [7] Method according to claim 6, characterized by that the procedure includes: Recording (S112) at least one extent, in particular an extent in a plane of the drivable surface (124) and / or a depth, of each detected damage site (122); wherein the generated data set contains all recorded extents of the damaged areas (122) and is configured to contain the associated at least one extent for each of the plurality of damaged areas (122). [8] Method according to one of the preceding claims, characterized by that procedure (V1) contains: Providing (S100) a classification scheme for classifying damaged areas (122) on the basis of at least one characteristic value of the damaged area (122), which characteristic value can be detected by the at least one sensor device (104, 106); Classifying (S114) each detected damage site (122) into a damage site class by applying the classification scheme to the at least one sensor signal; and wherein the generated data set contains the damage class for each classified damage location (122) and is configured to contain the associated damage class for each of the plurality of damage locations (122). [9] Method (V2) for autonomously parking a vehicle (100) based on a parking trajectory (120), comprising the steps: Providing (S200) a data set containing a parking trajectory (120) that can be followed by a vehicle (100), wherein the data set contains at least one damaged area position, wherein the damaged area position indicates a damaged area (122) of a region (126) of a drivable surface (124); Detecting (S204) an area of ​​a drivable surface by means of at least one sensor device; Checking (S206) the accessibility of the area, comprising the steps: Detecting (S208) the presence of a damaged area within the area of ​​the drivable surface based on a sensor signal from the sensor device; and Determining (S210) that the area is not passable if a distance between a position of a detected damage site and a damage site position contained in the data set is greater than a preset position threshold; autonomous driving (S228) of the vehicle (100) along the parking trajectory if the area in front of the vehicle is passable; and Otherwise, initiating (S232) an autonomous bypass (S236) of the detected damage site. [10] Method (V2) according to claim 9, characterized by that the data set provided is set up to contain the corresponding extent for each of the multiple damage sites; and that the check of the accessibility includes: Detecting (S212) an extent of the detected damage site (122) on the basis of the signal of the at least one sensor device (104, 106); and Determining (S214) that the area (126) in front of the vehicle is not passable if the detected extent is greater than the extent contained in the data set and associated with the damaged area (122) by more than a preset extent threshold. [11] Method (V2) according to one of claims 9 or 10, characterized by that the provided data set is configured to contain an associated damage class for each of the multiple damage sites (122); and that the method includes: Providing (S202) a classification scheme for classifying a damaged area (122) on the basis of at least one characteristic value of the damaged area (122), which characteristic value can be detected by the at least one sensor device (104, 106); where checking (S206) the navigability includes: Classifying (S216) the detected damage (122) into a damage class by applying the classification scheme to the at least one sensor signal; and Determining (S218) that the area (126) in front of the vehicle (100) is not passable if the damage class of the detected damage (122) is not equal to a damage class stored in the data set for the detected damage (122). [12] Method (V2) according to one of claims 9 to 11, characterized by that the procedure includes: Detecting (S230) the presence of a further damaged area (134) within a region (132) of the drivable surface (124) which is adjacent to a region (126) of the drivable surface (124) located in front of the vehicle along the parking trajectory (120), on the basis of a sensor signal from the sensor device (104, 106); whereby the autonomous bypassing (S236) of the further damaged area (134) is also initiated (S232). [13] Method (V2) according to one of claims 9 to 12, characterized by that the procedure includes: Detecting (S220) a non-existence of a damaged area (122) at a damaged area position contained in the data set; and Update (S222) the record based on the detected absence. [14] Method (V1) according to one of claims 9 to 13, characterized by that checking (S206) of the navigability includes: Detecting (S224) the presence of an obstacle (130) in, on and / or above the area (126) located in front of the vehicle (100) on the basis of a sensor signal from the at least one sensor device (104, 106); and Determining (S226) that the area (126) in front of the vehicle (100) is not passable if at least one obstacle (130) is present in the area (126) in front of the vehicle (100); wherein the autonomous bypassing (S236) of the obstacle (130) is initiated (S234). [15] Data set for autonomous parking of a vehicle (100) based on a parking trajectory (120), comprising: a parking trajectory (120) that can be followed by the vehicle (100); wherein the data set is configured to contain, for a plurality of damage points (122), an associated damage point position, as well as preferably at least one associated extent and / or an associated damage point class, wherein each damage point position indicates a damage point (122) of a region 126 of the drivable surface 124, which region (126) is located in front of the vehicle (100) at least during part of the journey when the vehicle (100) is traveling along the parking trajectory (120). [16] Computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method (V1, V2) according to one of claims 1 to 14. [17] Control device (102) for a vehicle (100), which is configured to carry out a method (V1, V2) according to one of claims 1 to 14. [18] Vehicle (100) containing: at least one sensor device (104, 106) arranged and configured to detect a surface (124) over which the vehicle (100) can travel; at least one actuator configured to specify a direction of travel and / or speed of the vehicle (100); and the control device (102) according to claim 17, which is connected in a communicable manner to the at least one sensor device (104, 106) and the at least one actuator.

Images