Computer-implemented method for behavior planning of a vehicle, processing device and vehicle control device

Abstract

The invention relates to a computer-implemented method (10) for behavior planning of a vehicle (14), at least comprising determining (16) at least one planned driving behavior (18) of the vehicle (14) depending on an environmental perception (20) of a vehicle environment (22) of the vehicle (14), determining (16) at least one spatially resolved degree of significance (34) of a functional impairment of the environmental perception (20) and a spatially resolved perceptual ability (36) of the environmental perception (20), each in relation to the planned driving behavior (18), determining (38) a risk (40) of the functional impairment of the environmental perception (20) on the planned driving behavior (18) at least depending on the degree of significance (34) and the perceptual ability (36), determining a risk (42) of the planned driving behavior (18) on the vehicle (14) and / or the vehicle environment (22) depending on the risk (40). the functional impairment,Calculating (44) the applicable driving behavior (46) depending on the planned driving behavior (18) and the risk (42) of the planned driving behavior. Furthermore, the invention relates to a processing device (12) and a vehicle control device (48).

Description

[0001] The invention relates to a computer-implemented method for behavior planning of a vehicle according to claim 1. Furthermore, the invention relates to a processing device and a vehicle control device. State of the art

[0002] Risk analysis for automated driving systems is of paramount importance for the operation of autonomous or automated vehicles. Risk analysis involves the systematic identification, assessment, and prioritization of risks that may arise from the vehicle's interaction with its environment. The risk assessment of a planned vehicle behavior on the vehicle and / or its environment is crucial. This risk assessment enables the early detection of potential hazards caused by planning errors.

[0003] US 2022 / 0315052 A1 describes a risk assessment of driving situations of an autonomous vehicle, in which the probability of dangerous traffic situations and the probability of relevant system errors are used to determine an estimated probability of occurrence of critical events for a certain period of time.

[0004] The risk of a driving situation or driving behavior depends on the severity, the probability of the risk event occurring, the controllability of the risk event, and the probability of system failure. The ISO 26262 standard plays a key role in defining safety requirements for electrical and / or electronic systems in vehicles, including the definition of safety integrity levels aimed at reducing risk to an acceptable level. The probability of system failure in the context of autonomous or automated driving systems requires a continuous assessment of the probability of errors in the environmental perception, planning, and actuation functions, according to ISO 21448 (SOTIF). This assessment must be dynamic and context-dependent to ensure the safety and reliability of the system under all operating conditions. Disclosure of the invention

[0005] According to the present invention, a computer-implemented method for behavior planning of a vehicle is proposed, comprising the features of claim 1. This allows the vehicle's behavior planning to be carried out depending on spatially resolved circumstances and conditions. The risk of functional impairments can be determined more accurately. The risk of the driving behavior to be implemented can be better assessed. Behavior planning can be carried out more appropriately to the situation. The driving safety of the vehicle is increased.

[0006] The vehicle can be a motor vehicle, a truck, a two-wheeled vehicle, a mobile robot, or an industrial robot. The vehicle can assume at least a semi-autonomous or autonomous operating state during driving.

[0007] The behavior planning may include planning a driving behavior of the vehicle, in particular at least one movement trajectory of the vehicle.

[0008] Perception of the environment may require detection of the environment. Perception of the environment may be performed by capturing and processing environmental sensor data from at least one environmental sensor of the vehicle. The environmental sensor may be a LIDAR sensor, a radar sensor, an ultrasonic sensor, and / or a camera. Perception of the environment may process and interpret the environmental sensor data provided by the at least one environmental sensor. Machine learning and image processing techniques may be used in perceiving the environment, in particular to recognize patterns in the environmental sensor data and / or to classify the at least one environmental object. Perception of the environment may include the creation of an environmental model of the vehicle's surroundings.

[0009] The environment perception can comprise at least one environment perception function. The environment perception function can include color recognition, object recognition, and / or object classification of at least one environment object.

[0010] Perceptual ability can be a performance capability, such as spatial resolution or recognition performance, of environmental perception. Perceptual ability can depend on the probability of a functional impairment in environmental perception.

[0011] The degree of significance of the functional impairment in relation to the planned driving behavior can indicate the extent of the functional impairment on the driving safety and / or vehicle safety of the vehicle in the planned driving behavior.

[0012] The risk of functional impairment on planned driving behavior can be proportional to the severity of the functional impairment. The risk of functional impairment can be inversely proportional to the perceptual ability to perceive the surroundings. The risk of functional impairment can also be spatially resolved.

[0013] The vehicle environment present during the planned driving behavior may include at least one environmental object. The environmental object may be a building, a living being, a device, an object, another vehicle, a plant, or the surface condition of a roadway on which the vehicle is traveling.

[0014] The planned driving behavior can include discrete parameters, such as trajectory, lane guidance, lane changes, and / or braking. The planned driving behavior can depend on the vehicle's operating parameters, in particular target speed, target point, acceleration, braking, acceleration gradient, braking gradient, steering angle, and / or steering angle gradient. The planned driving behavior can include at least one risk-minimized driving maneuver (MRM).

[0015] The planned driving behavior can be determined at specified time intervals, for example every 200 ms.

[0016] The functional impairment of environmental perception can be a limitation of performance, for example, a malfunction of environmental perception. The malfunction can, for example, consist of not detecting or recognizing an environmental object, detecting it too late, and / or incorrectly. The malfunction can be a malfunction of at least one environmental perception function of environmental perception. The malfunction can be a false positive detection, a false negative detection, an incorrect bounding box, an incorrect class, and / or an incorrect or delayed detection.

[0017] The functional impairment can occur randomly, for example due to electrical and / or electronic faults, and / or systematically. The functional impairment, especially one that occurs systematically, can arise depending on initiating events, for example internal and / or external influences, in particular the vehicle's environment. The external initiating events can include environmental conditions, in particular road conditions, weather conditions, and / or object states or object interactions of environmental objects. The internal initiating events can include system conditions, system states, or interactions within the vehicle's system, for example, computing power, storage capacity, transmission power, synchronization, latencies, electrical, electronic, and / or thermal states.

[0018] The degree of importance and / or the perceptual ability can be determined model-based, in particular through machine learning.

[0019] The risk of functional impairment can be determined based on the environmental perception function. The risk of functional impairment can be determined using models, particularly machine learning.

[0020] The determination of the risk of driving behavior can be carried out before, after or at the same time as the determination of the degree of significance, the ability to perceive and / or the risk of functional impairment.

[0021] In a preferred embodiment of the invention, it is advantageous if the degree of importance and / or the perceptual ability are determined depending on the driving situation and / or vehicle environment. The degree of importance and / or the perceptual ability can be determined depending on the at least one environment perception function of the environment perception.

[0022] The degree of importance, perceptibility and / or risk of functional impairment can be determined depending on environmental information, in particular map data, vehicle-to-vehicle (V2V) communication and / or information from infrastructure facilities.

[0023] In a preferred embodiment of the invention, it is provided that the degree of importance and / or the perception capability in spatial zones of the vehicle's surroundings is determined with spatial resolution. The risk of functional impairment with regard to an environment perception function can be higher in some spatial zones of the vehicle's surroundings than in others. For example, with color recognition as an environment perception function, the risk of functional impairment on the vehicle and / or the vehicle's surroundings can be higher in spatial zones enclosing a traffic light and possibly also spatial zones bordering the traffic light than in spatial zones outside thereof. With open space recognition as an environment perception function, the risk of functional impairment on the vehicle and / or the vehicle's surroundings can be higher in spatial zones immediately in front of the vehicle than in spatial zones outside thereof.When lateral motion detection is used as an environment perception function, the risk of functional impairment on the vehicle and / or the vehicle environment may be higher to the side of the vehicle than in spatial zones outside it.

[0024] The spatial zones can be divided by a grid structure, particularly a three-dimensional one.

[0025] In a specific embodiment of the invention, it is advantageous if the degree of importance and / or the perceptual ability are determined depending on the vehicle's environmental conditions. The environmental conditions can be weather conditions, visibility conditions, and / or road conditions.

[0026] In a preferred embodiment of the invention, it is provided that the risk of driving behavior associated with at least one risk event is further determined depending on a degree of significance of the risk event, a probability of occurrence of the risk event, and / or a controllability of the risk event. The risk event can comprise a violation, exceedance, and / or non-compliance with specifications, for example spatial references. The spatial references can be static or dynamic and / or constant or variable. The static and constant spatial references can be road markings and / or road boundaries. The static, variable spatial references can be traffic signs and / or traffic lights. The dynamic static spatial references can be distances to less dynamic road users, for example pedestrians.The dynamically changing spatial references can be distances to more dynamic road users, for example other vehicles.

[0027] The degree of significance of the risk event may depend on a mass and / or a speed of the vehicle and / or of the at least one environmental object associated with the risk event.

[0028] The probability of occurrence of the risk event may include a trajectory probability of road users and / or a braking probability of other vehicles.

[0029] Controllability can be influenced by the behavior of other road users.

[0030] In a preferred embodiment of the invention, it is advantageous if the risk of driving behavior is specified by at least one risk parameter. The risk parameter can be related to a spatial zone, a group of spatial zones, an environmental event, a sequence of environmental events, and / or a time period. The parameterization of the risk of driving behavior can be selected depending on the vehicle type.

[0031] The risk of functional impairment on the planned driving behavior can also be indicated by another risk parameter.

[0032] In a specific embodiment of the invention, it is advantageous if the applicable driving behavior deviates from the planned driving behavior if the risk parameter exceeds a threshold value. The applicable driving behavior can correspond to the planned driving behavior if the risk parameter at least falls below a threshold value. A reverse evaluation with respect to the risk parameter may also be possible.

[0033] In a preferred embodiment of the invention, it is advantageous if, in addition to the determined planned driving behavior, at least one planned second driving behavior is determined as the first driving behavior, for which the risk of functional impairment and the risk of the planned second driving behavior are determined analogously to the first driving behavior, and the calculation of the driving behavior to be used includes a selection from the first and second driving behavior depending on the risk of the respective planned driving behavior. The risk of the second driving behavior can be specified by at least one further risk parameter. The driving behavior to be used can be calculated depending on the risk parameter and the further risk parameter. The driving behavior to be used can be calculated depending on the risk parameter and the further risk parameter as well as the limit value.The planned first and second driving behavior can comprise at least one set of trajectories.

[0034] If the risk parameter and the additional risk parameter are both below the threshold value, the applicable driving behavior can be selected from the first and second driving behaviors based on the benefit or expediency for the vehicle and / or the vehicle's environment. If the risk parameter and the additional risk parameter are both above the threshold value, the applicable driving behavior can be selected based on the lowest value compared between the risk parameter and the additional risk parameter, in particular taking into account the expediency and / or benefit of the respective driving behavior for the vehicle and / or the vehicle's environment.

[0035] Furthermore, the invention relates to a non-transitory, computer-readable medium containing instructions, the execution of which causes at least one processor to carry out the computer-implemented method with at least one of the features described above.

[0036] According to the present invention, a processing device with the features of claim 9 is further proposed. The processing device may have at least one processor for at least partially executing the method.

[0037] According to the present invention, a vehicle control device with the features of claim 10 is further proposed. The vehicle control device can be a control unit or a part of a control unit of the vehicle. Setting the driving behavior to be applied can include initiating, controlling, and / or actuating the driving behavior to be applied. The vehicle control device can include the processing device.

[0038] Further advantages and advantageous embodiments of the invention emerge from the description of the figures and the illustrations. Character description

[0039] The invention is described in detail below with reference to the figures. They show in detail: Fig. 1: A computer-implemented method for behavior planning in a specific embodiment of the invention. Fig. 2 to 9: A respective driving situation when executing the method for behavior planning in a further specific embodiment of the invention.

[0040] Fig. 1 shows a computer-implemented method for behavior planning in a specific embodiment of the invention. The computer-implemented method 10 for behavior planning of a vehicle can be executed by a processing device 12 in the vehicle 14 and comprises at least determining 16 at least one planned driving behavior 18 of the vehicle 14 depending on an environmental perception 20 of a vehicle environment 22 of the vehicle 14. The environmental perception 20 may require an environmental detection 24 of environmental objects 26 in the vehicle environment 22 by at least one environmental sensor 30 of the vehicle 14.

[0041] The method further comprises determining 32 at least one spatially resolved degree of significance 34 of a functional impairment of the surroundings perception 20 and a spatially resolved perception capability 36 of the surroundings perception 20, each with respect to the planned driving behavior 18. The functional impairment of the surroundings perception 20 can be a limitation of performance, for example, a malfunction of the surroundings perception 20. The degree of significance 34 of the functional impairment with respect to the planned driving behavior 18 can indicate an extent of the functional impairment on driving safety and / or vehicle safety in the planned driving behavior 18. The perception capability 36 can be a performance, for example, a spatial resolution or a recognition performance, of the surroundings perception 20.

[0042] Furthermore, the method comprises determining 38 a risk 40 of the functional impairment of the environmental perception 20 on the planned driving behavior 18 at least depending on the degree of importance 34 and the perception ability 36, determining 41 a risk 42 of the planned driving behavior 18 on the vehicle 14 and / or the vehicle environment 22 depending on the risk 40 of the functional impairment and calculating 44 a driving behavior 46 to be applied depending on the planned driving behavior 18 and the risk 42 of the driving behavior.

[0043] A vehicle control device 48 of the vehicle 14 is configured to adjust the applicable driving behavior of the vehicle 14 depending on the calculated applicable driving behavior 46.

[0044] The degree of importance 34 and the perceptual ability 36 are determined in particular depending on the driving situation 50, the vehicle environment 22 and environmental conditions 54 of the vehicle 14, broken down into spatial zones 56 of the vehicle environment 22.

[0045] The risk 42 of driving behavior associated with at least one risk event 57 is further determined depending on a degree of significance 58 of the risk event, a probability of occurrence 60 of the risk event and a controllability 62 of the risk event.

[0046] The risk 42 of the driving behavior is specified by at least one risk parameter 64. The applicable driving behavior 46 is calculated differently from the planned driving behavior 18 if the risk parameter 64 exceeds or falls below a limit value 66.

[0047] In addition to the determined planned driving behavior 18 as the first driving behavior 68, at least one planned second driving behavior 70 is determined, for which the risk 40 of the functional impairment and the risk 42 of the planned second driving behavior 70 are determined analogously to the first driving behavior 68, and the calculation 44 of the driving behavior 46 to be applied includes a selection from the first and second driving behaviors 68, 70 depending on the risk 42 of the respective driving behavior.

[0048] Fig. 2 to shows a respective driving situation when executing the method for behavior planning in a further specific embodiment of the invention. In Fig. 2 shows the degree of significance 34 of the functional impairment of the surroundings perception, spatially resolved in spatial zones 56 of the vehicle environment 22, for three different surroundings perception functions in the same driving situation 50. The spatial zones 56 are designed in a grid structure, particularly three-dimensional, but shown here graphically as two-dimensional.

[0049] The illustrated classification of the degree of importance into a total of three levels L, a low level L1, a medium level L2 and a high level L3, applies to all figures unless otherwise stated.

[0050] In Fig. 2 a) shows color recognition as an environment perception function of the environment perception, in which the significance level 34 in the spatial zones 56 of a traffic light 72 assumes the high level L3, while the significance level 34 in the surrounding spatial zones 56 is, for example, zero. This means that a functional impairment, such as a malfunction, of the environment perception in the spatial zones 56 with the highest significance level 34 results in more serious consequences and thus a higher risk of the planned driving behavior than in the spatial zones 56 with a medium or low significance level L2, L1.

[0051] In Fig. 2 b) shows a free space detection as an environment perception function, in which the importance level 34 in the spatial zones 56 immediately in front of the vehicle 14 and between the vehicle 14 and a front vehicle 74 has the high level L3, further away therefrom the middle level L2 and around the front vehicle 74 the low level L1.

[0052] In Fig. 2 c) a lateral motion detection is shown as an environment perception function, in which the significance level 34 in the spatial zones 56 around a person 76 at the edge of the roadway 78 and in the spatial zones 56 to the side of the vehicle 14 has the high level L3.

[0053] In Fig. 3 shows the degree of significance 34 of the functional impairment, spatially resolved in spatial zones 56 of the vehicle environment 22, in a driving situation 50 in which a leading vehicle 74 is traveling in front of the vehicle 14 and in front of this, in turn, another leading vehicle 80. The degree of significance 34 in the spatial zones 56 behind the leading vehicle 74 has the high level L3, and in the spatial zones 56 behind the leading vehicle 74 traveling even further forward, the medium level L2. With distance detection as an environment perception function, a functional impairment in the spatial zones 56 with the high degree of significance 34 would have more serious consequences than in the spatial zones 56 with the medium degree of significance 34.

[0054] In Fig. 4 shows a driving situation 50 in which a person 76 is moving at the edge of the roadway in the direction of the roadway 78. The Fig. 4 a) shown driving situation 50 and vehicle environment 22 with the gradations of the degrees of importance 34 corresponds to the Fig. 2 c), while in Fig. 4 b) in the vehicle environment 22, a boundary 82, in particular a fence, is additionally present between the person 76 and the roadway 78. As a result, the severity level 34 of the functional impairment in the spatial zones 56 of the person 76 has the low level L1, so that even a malfunction of the environment perception in these spatial zones 56 would not have serious consequences in the interaction between the person 76 and the vehicle 14. The conditions in the vehicle environment 22, such as the boundary 82, can be obtained from map data.

[0055] In Fig. 5 shows a driving situation 50 and vehicle environment 22 in which a person 76 is at the edge of the road. Fig. 5 a) the person 76 moves further away from the roadway 78 and moves towards the roadway 78. The degree of significance 34 of the functional impairment depends on the driving situation 50 and the vehicle environment 22. The degree of significance 34 has the middle level L2 in the spatial zones 56 of the person 76 and since the person 76 as in Fig. 5 b) is getting closer and closer to the roadway 78, the degree of significance 34 of the functional impairment in the spatial zones 56 of the person 76 in time after the driving situation 50 in Fig. 5 a) the high level L3.

[0056] As in Fig. As shown in Figure 6, the perception capability 36 of the surroundings perception depends on the environmental conditions 54 of the vehicle environment 22. The perception capability 36 is also divided into three levels P, namely a low level P1, a medium level P2, and a high level P3. In Fig. 6 a) the perception ability 36 in good visibility and during the day in the spatial zones 56 of the vehicle in front 74 and the spatial zones 56 of the person 76 behind a tree 84 at the edge of the road has the lower level P1 than at night as environmental condition 54, as in Fig. 6 b). In poor visibility, for example in fog as the environmental condition 56, the perception capability 36 in the same spatial zones 56 is even higher, as in Fig. 6 c).

[0057] In Fig. 7 a) is the degree of significance 34 of the functional impairment, in Fig. 7 b) the ability to perceive 36 the environment and in Fig. 7 c) the risk 40 of impairment of the perception of the environment is depicted. As in Fig. 7 a), the severity level 34 of the functional impairment in the spatial zones 56 behind the front vehicle 74 and in the spatial zones 56 of the person 76 has the high level L3. As shown in Fig. As shown in Figure 7 b), the perceptual ability 36 in the spatial zones 56 around the vehicle 74 in front has the medium level P2 and behind a tree 84 at the edge of the road, behind which the person 76 is standing, the low level P1. Accordingly, the risk 40 of functional impairment, divided into three levels R, namely a low level R1, a medium level R2 and a high level R3, as shown in Fig. 7 c), as a function of the degree of importance and the ability to perceive, in the spatial zones 56 behind the front vehicle 74 and in the spatial zones 56 of the person 76 the high level R3.

[0058] In Fig. 8 a) is the degree of significance 34 of the functional impairment, in Fig. 8 b) a perceptual ability 36 of the environment perception and in Fig. 8 c) the risk 40 of impairment of the perception of the surroundings is depicted. As in Fig. As shown in Figure 8 a), the significance level 34 in the spatial zones 56 of the person 76 and the road edge has the high level L3. The perceptual ability 36 as in Fig. 8 b), has the medium level P2 in the spatial zones 56 around the front vehicle 74 and the low level P1 in the spatial zones 56 behind the tree 84 at the edge of the road, behind which the person 76 is standing. Accordingly, the risk 40 of functional impairment has the same level as in Fig. 8 c) shows, as a function of the degree of importance and the ability to perceive, the high level R3 in the spatial zones 56 of the person 76.

[0059] In Fig. 9 a) is the degree of significance 34 of the functional impairment, in Fig. 9 b) a perceptual ability 36 of the environment perception and in Fig. 9 c) the risk 40 of impaired function of the perception of surroundings is depicted. As in Fig. 9 a), the significance level 34 in the spatial zones 56 of the person 76 and the road edge as well as behind the front vehicle 74 has the middle level L2, since the vehicle 14 in comparison to Fig. 8 a) continues to the left. Also, the room zones 56 with the low level P1 of the Fig. 9 b) demonstrated perceptual ability 36 compared to those in Fig. 8 b) behind the tree 84. The risk 40 of functional impairment has, as in Fig. 9 c) shows, as a function of the degree of importance and the ability to perceive, at most the middle level R2. 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] US 2022 / 0315052 A1

[0003]

Claims

[1] Computer-implemented method (10) for behavior planning of a vehicle (14), at least comprising Determining (16) at least one planned driving behavior (18) of the vehicle (14) depending on an environmental perception (20) of a vehicle environment (22) of the vehicle (14), Determining (16) at least one spatially resolved degree of significance (34) of a functional impairment of the environmental perception (20) and a spatially resolved perceptual ability (36) of the environmental perception (20), each in relation to the planned driving behavior (18), Determining (38) a risk (40) of the functional impairment of the environmental perception (20) on the planned driving behavior (18) at least depending on the degree of importance (34) and the perception ability (36), Determining (41) a risk (42) of the planned driving behavior (18) on the vehicle (14) and / or the vehicle environment (22) depending on the risk (40) of the functional impairment, Calculation (44) of the driving behaviour (46) to be applied depending on the planned driving behaviour (18) and the risk (42) of the planned driving behaviour. [2] Computer-implemented method (10) for behavior planning according to claim 1, characterized by that the degree of importance (34) and / or the perceptual ability (36) is determined depending on the driving situation (50) and / or vehicle environment (22). [3] Computer-implemented method (10) for behavior planning according to claim 1 or 2, characterized by that the degree of importance (34) and / or the perceptual ability (36) in spatial zones (56) of the vehicle environment (22) is determined in a spatially resolved manner. [4] Computer-implemented method (10) for behavior planning according to one of the preceding claims, characterized by that the degree of importance (34) and / or the perceptual ability (36) is determined depending on the environmental conditions (54) of the vehicle (14). [5] Computer-implemented method (10) for behavior planning according to one of the preceding claims, characterized by that the risk (42) of the driving behavior associated with at least one risk event (57) is further determined depending on a degree of significance (58) of the risk event (57), a probability of occurrence (60) of the risk event (57) and / or a controllability (62) of the risk event (57). [6] Computer-implemented method (10) for behavior planning according to one of the preceding claims, characterized by that the risk (42) of the driving behavior is indicated by at least one risk parameter (64). [7] Computer-implemented method (10) for behavior planning according to claim 6, characterized by that the driving behaviour to be applied (46) is changed in a manner deviating from the planned driving behaviour (18) if the risk parameter (64) exceeds or falls below a limit value (66). [8] Computer-implemented method (10) for behavior planning according to one of the preceding claims, characterized by in that, in addition to the determined planned driving behavior (18), at least one planned second driving behavior (70) is determined as the first driving behavior (68), for which the risk (40) of functional impairment and the risk (42) of the planned second driving behavior (70) are determined analogously to the first driving behavior (68), and the calculation (44) of the driving behavior (46) to be used includes a selection from the first and second driving behaviors (68, 70) depending on the risk (42) of the respective planned driving behavior. [9] Processing device (12) for a vehicle (14) which is arranged to carry out the computer-implemented method (10) according to one of the preceding claims. [10] Vehicle control device (48) for a vehicle (14), which is designed to set an applicable driving behavior of the vehicle (14) depending on an applicable driving behavior (46) calculated using a computer-implemented method (10) according to one of claims 1 to 8.

Images