Method for determining criticality of evasion behavior of at least partially automated vehicle

Abstract

A general aspect of the present disclosure relates to a method for determining criticality of an evasion behavior of an at least partially automated vehicle advancing at a current speed. The method includes calculating a plurality of evasion trajectories of the vehicle relative to a reference collision object by combining different longitudinal accelerations and different lateral accelerations. The method further includes determining an optimal trajectory in which a minimum spacing between the vehicle and the collision object is maximum from the plurality of evasion trajectories. The method also includes determining a key acceleration vector for the associated key trajectory, wherein the minimum spacing reaches a limit range near the collision. The method further includes determining a criticality of the optimal trajectory based on the critical acceleration vector and an optimal acceleration vector belonging to the optimal trajectory.

Description

FIELD OF THE INVENTION

[0001] The present disclosure relates to driver assistance systems and methods having at least partially automated driving functions. In particular, a computer-implemented method for determining the criticality of an avoidance behavior of at least a partially automated vehicle is disclosed. BACKGROUND OF THE INVENTION

[0002] Driver assistance systems for driver warnings and for steering or braking interventions or corresponding driver warnings in dangerous situations are known. When further developing driver assistance systems into partially automated, automated or highly automated driving systems, various scenarios are sought to avoid collisions or mitigate the consequences of collisions. In principle, such driving systems can predict and / or execute automated avoidance maneuvers to avoid collisions, the avoidance maneuvers including pure braking, pure steering or combined braking and steering relative to a collision object detected, for example, by sensor data. By planning the trajectory of the vehicle relative to the collision object, it can be determined whether an avoidance maneuver is necessary and / or feasible. Here, it is very important in such driving systems to evaluate the avoidance behavior in terms of safety. In particular, such driving systems must be risk-proof (i.e., proven) before being put on the market. Safety assessment is also relevant for the optimal selection of avoidance maneuvers with respect to a minimum or at least reduced risk. For this purpose, various criticality indicators are provided, which enable a quantitative assessment of the driving behavior or avoidance maneuvers. Here, in principle, a distinction can be made between sub-critical situations in which a collision can be avoided and critical situations in which a collision is inevitable. In particular, the criticality indicator should reliably evaluate driving situations at low and high speeds.

[0003] The problem on which the present disclosure is based is how a quantitative assessment of the avoidance behavior of at least a partially automated vehicle can be provided more reliably. SUMMARY OF THE INVENTION

[0004] A first general aspect of the present disclosure relates to a computer-implemented method for determining the criticality of an avoidance behavior of an at least partially automated vehicle, where the at least partially automated vehicle is moving forward at a current speed. The method includes: calculating a plurality of avoidance trajectories of the vehicle relative to a reference collision object by combining different longitudinal accelerations and different lateral accelerations. The method further includes determining an optimal trajectory from the plurality of avoidance trajectories, in which the minimum distance between the vehicle and the collision object is maximized. The method also includes determining a critical acceleration vector of the associated critical trajectory, where the minimum distance reaches a limit range close to a collision. The method further includes determining the criticality of the optimal trajectory based on the critical acceleration vector and an optimal acceleration vector belonging to the optimal trajectory. In a design, determining the critical acceleration vector may include reducing the absolute value of the optimal acceleration vector to the critical acceleration vector where the minimum distance reaches a limit range close to a collision. In a design, the criticality of the optimal trajectory may be determined by forming a ratio of the absolute value of the critical acceleration vector to the absolute value of the optimal acceleration vector.

[0005] A second general aspect of the present disclosure relates to a computer system designed to execute a computer-implemented method for determining the criticality of an avoidance behavior of an at least partially automated vehicle according to the first general aspect (or an implementation thereof).

[0006] A third general aspect of the present disclosure relates to a computer program including instructions that, when the computer program is executed by a computer system, cause the computer system to implement a computer-implemented method for determining the criticality of an avoidance behavior of an at least partially automated vehicle according to the first general aspect (or an implementation thereof).

[0007] A fourth general aspect of the present disclosure relates to a computer-readable medium or signal that stores and / or contains a computer program according to the third general aspect (or an implementation thereof).

[0008] A fifth general aspect of the present disclosure relates to a method for testing and / or validating an at least partially automated driving function system. In this method, in order to enable the at least partially automated driving function system, the criticality is used as an evaluation criterion, and the criticality is determined by a computer-implemented method according to the first general aspect (or an implementation thereof).

[0009] A sixth general aspect of the present disclosure relates to a method for establishing an at least partially automated driving function system. In this method, the establishment of the at least partially automated driving function, in particular the establishment of rules for avoidance maneuvers, is performed at least based on the criticality determined by a computer-implemented method according to the first general aspect (or an implementation thereof).

[0010] The computer-implemented method according to the first general aspect (or an embodiment thereof) proposed in the present disclosure can be used to: obtain a quantitative assessment of the driving behavior (in particular, the avoidance behavior) of at least a partially automated vehicle. In particular, the computer-implemented method can be used to evaluate various avoidance maneuvers with respect to the maximum available vehicle capabilities. In other words, the evaluation can be performed by determining the criticality as proposed: that is, what share of the overall vehicle capabilities must be used to avoid a collision. This enables better comparability of criticalities in different driving situations, especially driving situations with different instantaneous speeds. In other words, a specific criticality has better comparability over a large speed range than the minimum distance. Thereby, the selection and protection of avoidance maneuvers regarding collision risk and / or regarding mitigation of collision consequences can be improved. Therefore, the partially automated driving function can be improved by the proposed method. Another advantage of the proposed computer-implemented method can be regarded as: reducing or increasing the intensity (e.g., braking intensity and / or steering intensity or longitudinal acceleration and / or lateral acceleration) of a driving maneuver (e.g., an avoidance maneuver) based on the quantitative knowledge of the share of the overall vehicle capabilities used. Thereby, advantages in terms of user satisfaction or reduced wear can be obtained again.

[0011] Some terms are used in the present disclosure as follows:

[0012] "Longitudinal acceleration" can be regarded as the absolute value of the acceleration in the longitudinal direction (i.e., in the longitudinal direction of the at least partially automated vehicle under consideration).

[0013] "Lateral acceleration" can be regarded as the absolute value of the acceleration in the lateral direction (i.e., in the lateral direction of the at least partially automated vehicle under consideration).

[0014] "Acceleration vector " is the combination of the longitudinal acceleration and the lateral acceleration.

[0015] "Vehicle" can be any device for transporting passengers and / or goods. The vehicle can be a motor vehicle (e.g., a car or a truck), but can also be a rail vehicle. The vehicle can also be a motorized two-wheeler or three-wheeler. Of course, floating and flying devices can also be vehicles. The vehicle can operate at least partially autonomously or assisted.

[0016] The term "at least partially automated" includes partially automated, automated, and / or highly automated devices and / or methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A method for determining the criticality of the avoidance behavior of at least a partially automated vehicle is schematically shown, as well as the use of the determined criticality in the following:

[0018] - In a method for testing and / or ensuring at least partially automated driving function systems, and

[0019] - In a method for establishing at least partially automated driving function systems.

[0020] Figure 2A Schematically shows the determination of an exemplary avoidance trajectory by combining longitudinal acceleration and lateral acceleration in a maximum friction model.

[0021] Figure 2B Schematically shows the determination of an exemplary avoidance trajectory by combining longitudinal acceleration and lateral acceleration in a maximum friction model.

[0022] Figure 3 Schematically shows an exemplary avoidance maneuver of at least partially automated vehicle, which can avoid a collision object with a spacing dmin according to the exemplary avoidance trajectory. Detailed Description

[0023] First, reference is made to Figure 1 、 Figure 2A and Figure 2B to discuss the technology of the present disclosure. Referring to Figure 3 , possible results and advantages obtained by a computer-implemented method 100 for determining the criticality KRIT of the avoidance behavior of at least partially automated vehicle 1 disclosed herein are discussed.

[0024] Figure 1 is a flowchart showing possible steps of a computer-implemented method 100 for determining the criticality KRIT of the avoidance behavior of at least partially automated vehicle 1, where vehicle 1 is shown at a current speed shown. The computer-implemented method 100 includes calculating 110 a plurality of avoidance trajectories τ of vehicle 1 relative to a reference collision object 3 by combining different longitudinal accelerations 12 and different lateral accelerations 14, 14a i . Further, in step 120, an optimal trajectory τ is determined from the plurality of avoidance trajectories τ i , in which the minimum spacing d between vehicle 1 and collision object 3 optimal is maximum. The method 100 further includes: determining 130 the critical acceleration vector of the critical trajectory τ optimal to reach the limit range of near collision. The method further includes: based on the critical acceleration vector min and the optimal acceleration vector belonging to the optimal trajectory τ krit and where the limit range of near collision is reached. The method further includes: based on the critical acceleration vector and the optimal acceleration vector belonging to the optimal trajectory τ optimal to determine the criticality KRIT of the avoidance behavior of vehicle 1. Determine the optimal trajectory τ optimal of criticality KRIT optimal .

[0025] In addition, as shown in Figure 1 , the present disclosure also relates to a method 200 for testing and / or ensuring at least a partially automated driving function system 2. In the method 200, in order to activate the at least partially automated driving function system 2, the criticality KRIT determined by the computer-implemented method 100 is used as an evaluation criterion.

[0026] In addition, referring to Figure 1 , the present disclosure also relates to a method 300 for establishing at least a partially automated driving function system 2. In the method 300, the at least partially automated driving function is established at least based on the criticality KRIT, which is determined by the computer-implemented method 100. In a design, the establishment of the at least partially automated driving function may include establishing rules for avoiding maneuvers. In a design, the control unit of the at least partially automated driving function system 2 may be established at least based on the criticality KRIT determined by the computer-implemented method 100. In a design, a backend system may be established, which is designed to communicate with the at least partially automated driving function system 2. In a design of the method 300, this establishment may include installing software in the memory of the at least partially automated driving function system 2. In a design, the software may be installed wirelessly.

[0027] Below, further designs of the computer-implemented method 100 are explained according to Figure 2A and 2B . Figure 2A and 2B show an exemplary method for determining an avoidance trajectory τ i or its associated acceleration vector . Figure 2A and 2BDescribe in detail the maximum friction model 10. It can be understood as the following model, which considers the force or acceleration that can be maximally transmitted from a vehicle to its surrounding environment. In the current example, the Kammscher circle model 10 familiar to those skilled in the art of road vehicles is used. Therefore, according to the vehicle and ground specifications, the range of the maximum transmissible force or acceleration can be schematically represented by the depicted circle. Here, the axes 12, 12a represent the longitudinal accelerations 12, 12a. Here, the axis 12 can start from the origin and upward represent the braking acceleration 12 or deceleration or backward acceleration. The axis 12a can start from the origin and downward represent the driving acceleration 12a or forward acceleration. Here, the axes 14, 14a represent the lateral accelerations 14, 14a, which can be caused, for example, by the steering of the vehicle 1. Here, the axis 14 can start from the origin and to the right characterize the lateral acceleration pointing to the right (for example, by turning left). The axis 14a can start from the origin and to the left characterize the lateral acceleration pointing to the left (for example, by turning right). The representation of the left - right acceleration and the front - back acceleration explained in this example can also be alternatively defined in other design solutions. What is important for the representation is the representation of the longitudinal acceleration or force orthogonal to the lateral acceleration or force.

[0028] According to Figure 2A the maximum friction model 10 shown in, the combinations included in step 110 of different longitudinal accelerations 12 and different lateral accelerations 14 can thus include combining the maximum available braking ability with the maximum available steering ability. In other words, the vehicle capabilities (i.e., the maximum transmissible frictional force) regarding steering and / or braking that are maximally available according to the maximum friction model 10 are determined. Different longitudinal accelerations 12 can be based on different braking intensities. Different lateral accelerations 14 can be based on different steering intensities. Here, the different ratios of the longitudinal acceleration 12 and the lateral acceleration 14 can be combined in such a way that different points are selected along the circle and the corresponding absolute values 12 i and 14 i . This is schematically shown in Figure 2A according to the curved arrow and the cross on the circle. Exemplarily, for the avoidance trajectory τ i the longitudinal acceleration 12 i and the lateral acceleration 14 i are determined. The synthesis of the longitudinal acceleration 12 i and the lateral acceleration 14 i forms the acceleration vector i belonging to the corresponding avoidance trajectory τ Via the trajectory calculation model known to those skilled in the art, the trajectory or the avoidance trajectory τ can be determined according to the current speed of the vehicle i . In other words, it can be based on the current speed Calculating a plurality of avoidance trajectories τ by combining different braking intensities and different steering intensities i . The calculation of the plurality of avoidance trajectories τ i may include, for example, calculating 2 to 1000, 10 to 500, or 50 to 100 avoidance trajectories τ i . In a design scenario, more than 1000 avoidance trajectories τ may also be calculated i . Even in the example shown in Figure 2A only the lateral acceleration 14 on one side is shown, the lateral acceleration 14a on the other side may also be considered in a design scenario. When combining different longitudinal accelerations 12 with different lateral accelerations 14, pure braking, pure steering, and values in between may be considered. In other words, by combining different longitudinal accelerations 12 with different lateral accelerations 14 according to the maximum friction force model 10, different maximum transmissible acceleration vectors can be formed from the braking vector 12 i and the steering vector 14 i as will be explained in detail below

[0029] As shown in Figure 3 , for each calculated avoidance trajectory τ i the minimum distance d between the vehicle 1 and the collision object 3 can be determined min . In a design scenario, the corresponding minimum distances d of the plurality of avoidance trajectories τ i can be compared with each other. Here, one or more avoidance trajectories τ in the plurality of avoidance trajectories τ min can be selected or determined that have the maximum minimum distance d i as the optimal avoidance trajectory τ min . i optimal opt .

[0030] In Figure 2B an exemplary optimal avoidance trajectory τ with the associated optimal acceleration vector and the corresponding optimal longitudinal acceleration 12 opt and optimal lateral acceleration 14 opt is shown optimal . As further shown in Figure 2B , determining the critical acceleration vector may include reducing the absolute value of the optimal acceleration vector to the critical acceleration vector where the minimum distance d min reaches the limit range close to the collision In a design scenario, as shown in Figure 2B , the absolute value can be gradually reduced, for example, by following the arrow of the acceleration vector in the direction towards the origin. The minimum distance d can be determined for each step sizemin . The stepwise reduction can be carried out in steps of the same size or of different sizes. In particular, the optimal acceleration vector can be iteratively reduced in absolute value until the minimum spacing dmin reaches the limit range close to a collision. In a design solution, the limit range close to a collision can be defined by d min > 0 and d min ≤ ε. Here, "ε" can be defined as the following proximity-to-collision spacing ε, in the case of which a collision between the vehicle 1 and the collision object 3 is just prevented. In a design solution, the proximity-to-collision spacing ε can be between 1 mm and 1000 mm, preferably between 10 mm and 500 mm, particularly preferably between 50 mm and 100 mm. In particular, the proximity-to-collision spacing ε can also be <1000 mm, <500 mm, <100 mm, <50 mm, <10 mm or <1 mm.

[0031] In a design solution, the criticality KRIT of the optimal trajectory τ can be determined by forming the ratio of the absolute value of the critical acceleration vector to the absolute value of the optimal acceleration vector optimal . That is, the following applies: optimal . Namely, it holds that:

[0032]

[0033] In particular, the proposed computer-implemented method is advantageous in the case of application in sub-critical situations, i.e., in situations where a collision can be avoided. Here, the value of KRIT can fluctuate between 0 and 1. The lower the value of the criticality KRIT, the lower the share of the maximum available vehicle capabilities required to prevent a collision. If and thus also the value of KRIT are equal to 0, no evasive maneuver is required. The greater the value of the criticality KRIT, the greater the share of the maximum available vehicle capabilities required to prevent a collision.

[0034] For the calculated evasive trajectory τ i for which no minimum spacing d min > 0 is produced, a collision cannot be prevented. That is, in such a driving situation, for the optimal trajectory τ optimal , d min is also negative. In a first variant of the computer-implemented method 100, in this case the criticality KRIT can be limited to 1. In a second variant of the computer-implemented method 100, in this case the criticality KRIT can also become greater than 1. Similar to the above reduction of the absolute value of the optimal acceleration vector , in the second variant, the optimal acceleration vector The absolute value can be increased until the critical acceleration vector In the case of the critical acceleration vector the critical trajectory (τ krit ) is determined, and in the case of the critical trajectory (τ krit ) the minimum spacing (d min ) again reaches the limiting range close to a collision. Thus, a risk assessment can also be carried out for critical situations where a collision cannot be avoided, based on the maximum available vehicle capabilities.

[0035] In a design of the computer-implemented method 100, the associated acceleration vectors i can be determined for a plurality, in particular all, of the calculated avoidance trajectories τ critical acceleration vectors and the corresponding criticality KRIT i . Furthermore, an optimized trajectory τ optimal can be determined based on a comparison of specific criticalities KRIT i . The trajectory τ optimiert with the lowest critical value can be determined as the optimized trajectory τ i optimiert . For a specific criticality KRIT, an explanation similar to the above can apply:

[0036]

[0037] In a design, the computer-implemented method 100 can include action steps for displaying the optimal trajectory τ optimal and / or its criticality KRIT optimal . In a design, the display can include an acoustic output, an optical output, and / or a vibration output. In a design, the computer-implemented method 100 can include action steps for autonomously or semi-autonomously adjusting the avoidance maneuver of at least partially automated vehicle 1 according to the optimal trajectory τ optimal . In a design, specific criticalities KRIT, KRIT i , KRIT optimal can be used to test and / or guarantee at least the at least partially automated driving function system 2 of at least partially automated vehicle 1. For example, activation can be refused starting from a specific criticality value (e.g., from KRIT = 0.7 or greater).

[0038] In this regard, Figure 3 ​Schematically shows at least partially automated vehicle 1 and reference collision object 3. The at least partially automated vehicle 1 can be equipped with an at least partially automated driving function system 2. The at least partially automated driving function system 2 can include a sensor system and / or obtain sensor data from the sensor system. The sensor system can be designed to: detect vehicle data and environmental data. For example, the sensor system can include long-range and mid-range environmental sensors, such as based on long-range radar (LRR), lidar, mid-range radar (MRR), short-range radar (SRR), ultrasonic, video, etc. Information about the reference collision object 3 can be obtained via the sensor system. The at least partially automated driving function system 2 can be designed to execute the above computer-implemented method 100. The at least partially automated vehicle 1 is at a current speed moving in the longitudinal direction (along arrow ). Exemplarily, the calculated avoidance trajectory τ Figure 3 is shown in i , and a collision with the collision object 3 can be prevented by means of the avoidance trajectory. It goes without saying that the collision object 3 or its position can be observed statically or dynamically. Various schemes for dynamically observing the collision object are known to those skilled in the art. For example, the position or movement of the collision object 3 (i.e., the change in position) can be projected into the future. The projection of the movement of the collision object 3 can be performed, for example, with the aid of a constant velocity vector. In other examples, the projection of the movement of the collision object 3 can include a change in the velocity vector of the collision object 3 (e.g., by a change in the direction and / or absolute value of the velocity vector), in particular a change in the direction towards the trajectory of the vehicle 1. Additionally, the minimum distance d i belonging to the exemplary avoidance trajectory τ min is shown. As described above, the computer-implemented method 100 can be executed during the operation of the at least partially automated vehicle 1. For example, the shown avoidance trajectory τ i can correspond to the optimal trajectory τ optimal . The computer-implemented method 100 can cause the at least partially automated vehicle 1 to continue to be controlled by outputting steering and / or braking signals corresponding to the optimal trajectory τ optimal . Alternatively or additionally, warnings and / or recommendations can be made via the user interface. Alternatively or additionally, the method 100 can be used to test and / or guarantee the at least partially automated driving function system 2 of the at least partially automated vehicle 1.

[0039] Furthermore, a computer system is disclosed, which is designed to: execute for determining at a current speed Method 100 for determining the criticality KRIT of an avoidance maneuver of at least partially automated vehicle 1 moving forward. The computer system can include at least one processor and / or at least one working memory. The computer system can also include (non-volatile) memory. In an example, all steps of method 100 can be executed by the computer system. In some examples, the individual steps of method 100 can be executed by the computer system. Optionally, the computer system can receive the results of the individual method steps that are not executed by the computer system. In a design, the computer system can be included in the at least partially automated driving function system 2 described above.

[0040] There is also disclosed a computer program designed to execute a method 100 for determining the criticality KRIT of an avoidance maneuver of at least partially automated vehicle 1 moving forward at a current speed Method 100 for determining the criticality KRIT of an avoidance maneuver of at least partially automated vehicle 1 moving forward. For example, the computer program can exist in an interpretable or compiled form. The computer program can be (even partially) loaded into the RAM of the computer for execution, for example as a sequence of bits or bytes. In a design, the computer program can be installed in the memory of the at least partially automated driving function system 2 described above.

[0041] There is also disclosed a computer-readable medium or signal storing and / or containing the computer program or at least a part thereof. The medium can include, for example, one of RAM, ROM, EPROM, HDD, SDD... on / wherein the signal is stored.

Claims

1. A computer-implemented method (100) for determining a criticality (KRIT) of an evasive maneuver of an at least partially automated vehicle (1), wherein the at least partially automated vehicle (1) is moving at a current speed The method (100) comprises the following steps: - calculating (110) a plurality of evasive trajectories (τ) of the vehicle (1) relative to a reference collision object (3) by combining different longitudinal accelerations (12) and different lateral accelerations (14) i ); - from the plurality of avoidance trajectories (τ i ) to determine the optimal trajectory (τ) in (120) optimal ), the minimum distance (d ) between the vehicle and the collision object (3) in the case of the optimal trajectory min )maximum; - Determine the critical trajectory (τ) to which (130) belongs krit )'s key acceleration vector In the case of the critical trajectory, the minimum distance (d min ) reaches the limit range close to collision; -Based on the key acceleration vector and belongs to the optimal trajectory (τ optimal )'s optimal acceleration vector Determine (140) the optimal trajectory (τ optimal )'s criticality (KRIT optimal ).

2. The computer-implemented method (100) of claim 1, wherein determining the key acceleration vector Including the optimal acceleration vector The absolute value of the minimum spacing (d min ) The critical acceleration vector that reaches the limit of the collision 3. The computer-implemented method (100) according to any one of the preceding claims, wherein the key acceleration vector The absolute value of the optimal acceleration vector The optimal trajectory (τ optimal )’s criticality (KRIT).

4. The computer-implemented method (100) according to any one of the preceding claims, wherein the limit range of the near-collision is determined by (d min >0) and (d min ≤ε), where (ε) defines the near collision spacing that just prevents collision.

5. A computer-implemented method (100) according to any of the preceding claims, wherein combining different longitudinal accelerations (12) with different lateral accelerations (14) includes combining a maximum available braking capacity and a maximum available steering capacity corresponding to a preset maximum friction model, in particular a preset Kammschen circle model (10).

6. The computer-implemented method (100) according to any one of the preceding claims, wherein for each avoidance trajectory (τ i ) determines a minimum distance (d ) between the vehicle (1) and the collision object (3) min ), and optionally, The multiple avoidance trajectories (τ i ) in the corresponding minimum spacing (d min ) are compared with each other, and wherein the plurality of avoidance trajectories (τ i ) has the largest minimum spacing (d min )’s avoidance trajectory (τ optimal ) is selected as the optimal avoidance trajectory (τ optimal ).

7. The computer-implemented method (100) according to any one of the preceding claims, wherein for a plurality, in particular for all, of the calculated plurality of avoidance trajectories (τi) an associated acceleration vector is determined Key acceleration vector and the corresponding criticality (KRIT i ), and optionally - Based on the determined criticality (KRIT optimal ,KRIT i ) to determine the optimization trajectory (τ optimiert ).

8. The computer-implemented method (100) according to any one of the preceding claims, further comprising the following steps: - Display the optimal trajectory (τ optimal ) and / or the criticality of the optimal trajectory (KRIT optimal ), and / or - According to the optimal trajectory (τ optimal ) autonomously or semi-autonomously adjust the evasive maneuver of the at least partially automated vehicle (1).

9. The computer-implemented method (100) according to any one of the preceding claims, wherein the specific criticality (KRIT i ,KRIT optimal ) is used to test and / or ensure an at least partially automated driving function system (2) of the at least partially automated vehicle (1).

10. A computer system designed to carry out the method (100) for determining a criticality (KRIT) of an evasive manoeuvre of an at least partially automated vehicle (1) according to any one of the preceding claims.

11. A computer program comprising instructions which, when the computer program is executed by a computer system, cause the computer system to implement a method (100) for determining a criticality (KRIT) of an evasive maneuver of an at least partially automated vehicle (1) according to any one of claims 1 to 9.

12. A computer readable medium or signal storing and / or containing a computer program according to claim 11.

13. A method (300) for establishing an at least partially automated driving function system (2), wherein a criticality (KRIT) determined according to a computer-implemented method (100) according to any one of the preceding claims 1 to 9 is taken into account when establishing the at least partially automated driving function, in particular when establishing rules for evasive maneuvers.

Images