Method for determining an optimal stopping position for a vehicle, assistance system and vehicle

Abstract

The present invention relates to a method for determining an optimal stopping position for a vehicle (1), wherein the vehicle (1) has at least one ambient environment sensor that detects the ambient environment, and performs ambient environment analysis or ambient environment interpretation based on the detected ambient environment to determine a traffic situation, and when the vehicle (1) stops in front of an obstacle (13) due to the traffic situation, an optimal stopping position in front of the detected obstacle (13) is determined, and a grid map (16) including a plurality of grid cells is created to determine the optimal stopping position, the grid cells correspond to the determined stopping positions and quality levels of the stopping positions, and the optimal stopping position is selected from the stopping positions on the grid map (16) based on the quality level of each stopping position.

Description

[Technical Field]

[0001] The invention relates to a method for determining an optimum stopping position for a vehicle, and to an assistance system for determining a stopping position for a vehicle according to the method according to the invention, as well as to a vehicle having an assistance system according to the invention. [Background technology]

[0002] General-purpose vehicles, such as cars, trucks, and motorcycles, are increasingly equipped with assistance or driver assistance systems that, with the aid of sensor systems, detect the surrounding environment, recognize traffic situations, and assist the driver, for example, by braking or steering interventions or by outputting optical, tactile, or acoustic warnings. Radar sensors, lidar sensors, camera sensors, ultrasonic sensors, etc., are commonly used as sensor systems for detecting the surrounding environment. Inferences about the surrounding environment can then be drawn from the sensor data determined by the sensors, for example, so-called surrounding environment models can be generated. Based on these, warnings / information can be provided to the driver, or adjusted steering, braking, or acceleration instructions can be output. Thus, assistance functions that process sensor or surrounding environment data can, for example, assist or completely replace (semi-automatically or fully automatically) the driving task or vehicle control, thereby avoiding accidents with other road users or facilitating complex driving maneuvers. For example, a vehicle may perform autonomous emergency braking (AEB, Automatic Emergency Braking) with the assistance of an emergency braking assist (EBA, Emergency Braking Assist), speed adjustments and subsequent driving adjustments with the assistance of an adaptive cruise control assist (ACC), or may maintain the vehicle in its lane with a steering assist (LKA, Lane Keeping Assist).Here, there are assistance functions that are typically activated automatically or that initiate an automatic intervention, such as EBA, or assistance functions that are typically activated by the driver, such as ACC.

[0003] Autonomous vehicles in urban traffic can manage various maneuvers, such as "stay within a lane," "change lanes," "stop at a traffic light," and "stop behind an object." Here, a planning level is separated, with a maneuver planner defining basic maneuvers such as "stop," "change lanes," and "drive at speed x." Below this level is a trajectory planner, which calculates trajectories (e.g., x / y point lists with speed information) for higher-resolution maneuvers. However, even with the maneuver planner, there are situations where sufficient or minimal road safety is ensured, especially for detected obstacles. For example, if an autonomous vehicle stops behind an obstacle, if it stops too close to the obstacle, the vehicle's field of view (FOV) may be too narrow to recognize oncoming traffic early enough to resume driving. Or, if it stops too far from the obstacle to be recognized as "automated but natural driving style," the vehicle may stop. Therefore, we are particularly interested in finding an optimal stopping position behind an obstacle that guarantees a good compromise between "good visibility of the planned trajectory", "stopping close to the object" and "good starting course" (comfort, high performance). Summary of the Invention [Problem to be solved by the invention]

[0004] A 360° threat recognition sensor system for a vehicle capable of monitoring a focus area all around the vehicle is known from U.S. Pat. No. 11,226,624 B2. Here, a computer selects a focus area around a vehicle stopping position to be monitored by the sensor system. If the field of view of the focus area of ​​the sensor system is limited, the computer determines a new vehicle stopping position at a position where the field of view of the focus area of ​​the sensor system is not limited when the vehicle stops at the new vehicle stopping position. The computer then outputs the new vehicle stopping position to a vehicle occupant or sends a command to autonomously control the vehicle to stop at the new stopping position.

[0005] SUMMARY OF THE INVENTION Moving forward from the prior art, it is an object of the present invention to provide a method for determining an optimal stopping position behind a detected obstacle. [Means for solving the problem]

[0006] The above-mentioned object is achieved by the full content of claim 1 and the other independent claims. Preferred designs of the invention are set out in the dependent claims.

[0007] In the method for determining an optimal stopping position for a vehicle according to the present invention, the vehicle has at least one ambient environment sensor for detecting the ambient environment, and ambient environment analysis and / or ambient environment interpretation are performed to determine a traffic situation based on the detected ambient environment. As soon as it is assumed that the vehicle will stop in front of an obstacle due to the traffic situation, an optimal stopping position in front of the detected obstacle is then determined. In a next step, a grid map is prepared containing a number of grid cells for determining the optimal stopping position, the grid cells corresponding to the determined stopping positions and the quality of these stopping positions, and an optimal stopping position is selected from the stopping positions of the grid map based on the quality of each stopping position.

[0008] The driving planner advantageously optimizes the stopping position behind the obstacle to ensure a smooth departure with maximum visibility in areas obscured by allowing oncoming traffic to pass (natural driving behavior). The visibility area of ​​the oncoming road is optimized in this way. This reduces the risk of sudden braking to avoid a collision with an oncoming vehicle, for example, while at the same time increasing the safety of the system and driver / occupant acceptance, since assertive automated driving (not overly cautious automated driving behavior) occurs and sudden braking is also necessarily included in the range of traditional driving behavior to which occupants are accustomed. The resulting driving behavior increasingly resembles the driving behavior of an average human, who, in a corresponding driving situation where the vehicle needs to change lanes to go around an obstacle and stop at a certain distance behind the obstacle, not only stops as close to the obstacle as possible, but also drives along the planned road for as long as possible, ensuring that the target lane or adjacent road is visible from the stopping position.

[0009] A trajectory of the vehicle is preferably determined which is intended to allow the vehicle to avoid obstacles when following the trajectory.

[0010] A trajectory intended to avoid the obstacle when the vehicle follows the trajectory can be quickly determined. Then, taking into account or in relation to the determined (avoidance) trajectory, an attention area can be determined, which is located, for example, in a target lane or road for driving near the obstacle where oncoming traffic can or needs to be anticipated.

[0011] Furthermore, the field of view and / or line of sight of at least one ambient environment sensor or multiple ambient environment sensors (oriented in one direction) can be determined, thus determining variables that can be used to determine or calculate quality, for example.

[0012] Additionally, an overlap region between the region of interest and the field of view can be determined, describing the region where the region of interest and the field of view intersect. Advantageously, a larger overlap region is desirable to ensure the best possible sensor field of view, and therefore the most advantageous sensor data detection.

[0013] The trajectory and / or field of view and / or area of ​​interest and / or overlap area can be quickly determined for one or each grid cell, and for each grid cell or each stopping position, the trajectory, field of view, area of ​​interest and / or overlap area are determined to generate a statement based on these as to which quality each grid cell or each stopping position corresponds to.

[0014] The quality of each stopping position is therefore preferably defined based on the following (non-exhaustive) parameters: distance to obstacles, and / or size of the field of view, and / or size of the overlapping area, and / or size or length of the line of sight, or the path and / or trajectory path (regarding both the starting position of the trajectory, i.e., assessment of whether it is easy or difficult to travel along the trajectory, and the type of vehicle control depending on whether the trajectory extends in a favorable straight or unfavorable curved manner). Thus, for example, the path of the trajectory leading to the stopping position and / or the path of the trajectory leaving the stopping position, and / or the comfort of the trajectory leading to the stopping position and / or the comfort of the trajectory leaving the stopping position, and / or the safety of the trajectory leading to the stopping position and / or the safety of the trajectory leaving the stopping position, and / or the drivability of the trajectory leaving the stopping position, and / or the distance of the stopping position from unfavorable areas and / or the distance of the stopping position from favorable areas can be used to define the quality of each stopping position.

[0015] The cost of at least one parameter can be quickly determined based on a cost function, and / or the cost of at least one trajectory can be determined, which can be used to determine quality. In a practical manner, each determined trajectory can be first assessed from a stop position or grid cell using a cost function.

[0016] The present invention further claims a vehicle assistance system including a control unit for controlling a vehicle and at least one sensor for detecting the surrounding environment, wherein an environment analysis and / or an environment interpretation is performed based on the detected surrounding environment to determine the traffic situation. Furthermore, if the currently existing traffic situation requires the vehicle to stop in front of an obstacle, an optimal stopping position for the vehicle in front of the detected obstacle is determined based on the method according to the present invention.

[0017] Furthermore, the invention includes a vehicle having an assistance system according to the invention.

[0018] In the context of the environment analysis or interpretation, for example, objects detected by environment sensors are identified and their movement parameters (speed, acceleration, direction of movement, etc.) are determined and possibly tracked. Furthermore, an object classification procedure (so-called object classification) can be performed, in which the objects are classified into different object classes (different road users, etc.) based on their properties. The respective traffic situation present is then determined from the collected data, and the detected objects and their associated parameters are used in a specific way to determine or interpret the traffic situation. Furthermore, data on the detected objects can be summarized in an object list and / or an environment map. Furthermore, further information, for example from other road users (vehicle-to-vehicle communication) and / or infrastructure (vehicle-to-X communication) and / or navigation data, can also be used for the determination and interpretation. As an extension, for this purpose, a so-called environment model can be prepared based on this data, on the basis of which the vehicle can be controlled autonomously or at least semi-autonomously.

[0019] The invention is explained in more detail below on the basis of useful exemplary embodiments. [Brief explanation of the drawings]

[0020] [Figure 1] 1 shows a simplified schematic representation of an ego vehicle with an assistance system according to the invention; [Figure 2] A simplified representation of a traffic situation is shown in which a vehicle detects an obstacle in its own lane and determines an avoidance trajectory to go around the obstacle when there is an oncoming vehicle in an adjacent lane. [Figure 3] A simplified representation of the traffic situation in Figure 2 is shown, where a vehicle occupies a stopping position very close to an obstacle. [Figure 4] 3 shows a simplified representation of the traffic situation in Figure 2 where a vehicle occupies a stopping position at a distance from an obstacle. [Figure 5] 3 shows a simplified representation of the traffic situation of FIG. 2 in which vehicles occupy optimal stopping positions determined according to the method of the present invention. [Figure 6]1 shows a simplified representation of a traffic situation with vehicles occupying various stopping positions diagrammatically. [Figure 7] 1 shows a simplified representation of the design of grid map decidability in accordance with the present invention. [Figure 8] 1 shows a simplified representation of an undesirable scenario in which an oncoming vehicle appears at the top edge of the region of interest as the vehicle begins to move away from a stopped position. [Figure 9A] 9 shows a simplified graphical representation of the evolution of the speed of the ego vehicle in the traffic situation of FIG. 8. [Figure 9B] 9 shows a simplified representation of the evolution of the speed of oncoming vehicles in the traffic situation of FIG. 8. [Figure 10] 1 shows a simplified representation of a further traffic situation that first determines the trajectory for a vehicle to reach an optimal stopping position. [Figure 11] A simplified representation of the traffic situation in Figure 10 shows a vehicle starting from an optimal stopping position and now following an avoidance trajectory. [Figure 12] 1 shows a simplified representation of a further traffic situation in which a vehicle detects an obstacle in its own lane, there are no oncoming vehicles in adjacent lanes, and it determines a trajectory to avoid the obstacle. [Figure 13] 1 shows a simplified schematic representation of the design of the system architecture of the assistance system according to the invention; [Figure 14] 1 shows a simplified schematic representation of the design of the operation planning part of the assistance system according to the invention. [Figure 15] 1 shows a simplified schematic representation of a selection matrix design that can be used to create a grid map. DETAILED DESCRIPTION OF THE INVENTION

[0021] Reference numeral 1 in FIG. 1 indicates a host vehicle, or a vehicle having various actuators (steering 3, motor 4, brake 5) with a control unit 2 (ECU (electronic control unit) or ADCU (driver assistance / autonomous driving control unit)), and in particular, the control unit 2 is able to access the actuators, thereby enabling (semi-)automatic control of the vehicle 1. Furthermore, the vehicle 1 has sensors for detecting the surrounding environment, i.e., surrounding environment sensors (forward camera or cameras 6, lidar sensor 7, radar sensors 8 (long-range radar sensor, LRR) or 9a-9d (short-range radar sensors, SRR), ultrasonic sensors (USS) 10a-10d, and surround-view cameras 11a-11d), and converts these sensor data into the surrounding environment. By using the sensors for environment or object recognition, (semi-)autonomous vehicle behavior and various assistance functions can be realized, such as parking assistance, emergency braking assistance (EBA, Electronic Brake Assist), adaptive cruise control (ACC), lane keeping control or lane keeping assist (LKA, Lane Keeping Assist). The assistance functions are here executed via the control unit 2 or algorithms stored therein. Furthermore, further subordinate control units (ECUs; Electronic Control Units) can also be provided, for example to activate a surround view system. Furthermore, different arrangements of sensors are also included within the scope of the present invention, for example, four or more radar sensors, surround view cameras, ultrasonic sensors can also be provided, which can be arranged anywhere on the vehicle 1.

[0022] FIG. 2 illustrates a traffic situation in which vehicle 1 or the host vehicle is traveling on a road, an oncoming vehicle 12 is located in an adjacent lane of the road, and an obstacle 13 (simplified as a rectangle) is located in the lane in the vehicle 1's direction of travel. Vehicle 1 detects the obstacle 13 with the aid of sensors in its surrounding environment (cameras 6, 11a-11d, lidar 7, radars 8, 9a-9d, ultrasound 10a-10d, etc.) and calculates a corresponding or avoidance trajectory intended to avoid the obstacle 13 (shown in FIG. 2 based on the black arrow). Due to the trajectory, vehicle 1 executes a lane change into an adjacent lane, but an oncoming vehicle 12 is present in the adjacent lane. Based on the surrounding environment data and situation analysis or situation interpretation, vehicle 1 can now establish that changing into the adjacent lane will result in a collision with the oncoming vehicle 12. Based on this traffic situation, vehicle 1 determines a stopping position ahead of the obstacle 13 at which vehicle 1 will stop. In this case, it is particularly decided to find an optimal stopping position, which generally represents a compromise between various approaches or parameters. On the one hand, it should be ensured that the vehicle 1 stops close enough to the obstacle 13 so as not to unnecessarily disrupt the traffic flow. Furthermore, the vehicle 1 stops at a sufficient distance from the obstacle 13 to ensure, on the one hand, that the obstacle 13 is far enough away so that the vehicle 1 has enough space to traverse the avoidance trajectory as comfortably and safely as possible, and that the surrounding environment sensors have a good view (field of view of the surrounding environment sensors). Therefore, in addition to a stopping position close to the obstacle 13, good visibility for the planned trajectory is a priority, and consideration should be given to which stopping position provides the most favorable starting point, i.e. the start of the avoidance trajectory.

[0023] FIG. 3 illustrates the traffic situation of FIG. 2 together with the field of view 14 of the vehicle 1's surrounding environment sensor (for clarity, FIG. 3 only shows the field of view of the forward-facing surrounding environment sensor or radar sensor). It also shows a region of interest 15 (ROI), which indicates the area of ​​the adjacent lane that is of particular interest for the calculation of the avoidance trajectory. The region of interest 15 is, for example, summed with a definable standard value (which may correspond, for example, to the lane width and may be several meters) from the vehicle 1 in the adjacent lane, and its forward and opposite boundaries are defined by the current scenario (e.g., avoiding an oncoming road in a city center with a speed limit of 50 km / h). FIG. 3 shows that in this case, the vehicle 1 stops in an unfavorable position because the field of view 14 only covers or does not overlap a relatively small portion of the region of interest 15. The area of ​​the region of interest 15 that overlaps with the field of view 14 is also referred to as the overlap area. The overlap area in FIG. 3 is insufficient for the vehicle 12 to detect oncoming traffic, particularly because the sensor's field of view is limited by the obstacle 13.

[0024] In contrast, an optimized stopping position for vehicle 1 compared to the stopping position of Figure 3 is shown in Figure 4. In this case, the significantly greater distance between vehicle 1 and obstacle 13 results in better visibility or sensor field of view, a larger overlap area between field of view 14 and area of ​​interest 15, and sufficient reliability in detecting oncoming traffic. Nevertheless, the stopping position currently occupied by vehicle 1 is one that can be optimized to stop vehicle 1 closer to obstacle 13, e.g., to avoid unnecessarily restricting traffic flow.

[0025] FIG. 5 shows a traffic situation in which a vehicle 1 determines an optimal stopping position with the aid of the method according to the invention. Advantageously, the stopping position thus determined represents a compromise between good sensor visibility, a good starting position for the planned trajectory, and a sufficiently short distance to the obstacle 13. For this purpose, a grid map 16 is provided, which comprises a number of grid cells, each of which corresponds to a determined stopping position and a quality associated with each stopping position. In this case, the grid cells are shown as squares, but other shapes (e.g., rectangular, polygonal, circular, etc.) are also conceivable, provided they are arranged in a grid. FIG. 5 shows the quality of each grid cell based on its pattern: black grid cells represent low quality, striped grid cells represent medium quality, and white grid cells represent high quality. Here, quality was determined by observing for each grid cell or each stop position which has the most comfortable or straightest possible path in terms of visibility in the direction of the planned trajectory (maximum possible visual range), distance to the obstacle 13 (as close as possible to the object), and starting position of the trajectory in terms of drivability (within the limits of the driving dynamics of the vehicle 1). Here, individual parameters can be weighted differently and calculated accordingly. For example, a black grid cell indicates a stop position close to the obstacle 13, but with a low visual range of the sensor, representing a starting position that is relatively unfavorable for the trajectory to be driven (and therefore of low quality) due to its proximity to the obstacle 13. In contrast, a white grid cell indicates a stop position with an acceptable distance to the obstacle 13, good sensor visibility, and a good starting position (and therefore of high quality).

[0026] FIG. 6 shows two examples of different possibilities that can be used to determine each stopping position based on the vehicles 1A and 1B corresponding to vehicle 1. First, the determined attention area 15 is used as input for the grid map or individual stopping positions. The size of the attention area 15 is calculated to reduce the potential risk due to oncoming objects in adjacent lanes that would likely not be detected outside the attention area 15. Furthermore, the overlap area is assessed in terms of its size, i.e., the size of the area where the attention area 15 and the field of view 14 overlap, with the largest possible overlap area being advantageous. The system is designed so that a sufficiently large attention area 15 with a large overlap area reduces the risk of a collision due to a wrong starting decision by vehicle 1. If vehicle 1 starts on an oncoming road and then an object or vehicle 12 appears at the edge of the attention area 15, a collision can be prevented by comfortable braking actions by both vehicles. The feasibility of a trajectory starting from a grid cell is first checked. If no trajectory is found within the driving dynamics constraints (e.g., maximum steering angle and maximum steering angular velocity), the grid cell is discarded. This is the case, for example, for the black grid cell immediately preceding the obstacle on the right side of Figure 7.

[0027] Also, a cost function of the trajectory can be determined, starting from a starting position and using driving dynamics parameters (steering angle, steering angular velocity, acceleration, jerk, path angle, etc.) and the distance (safety distance) to the obstacle 13. A further parameter of the cost function may be the distance traveled by the vehicle 1 along the trajectory after a certain time has elapsed.

[0028] The results showed that the trajectory, and therefore the stopping position based on the cost function of the trajectory of vehicle 1B, was lower and more advantageous than the trajectory and stopping position of vehicle 1A.

[0029] FIG. 7 shows an example of calculations of how the individual stop positions or trajectories and quality of each grid cell can be determined or calculated. In FIG. 7, the line of sight of the sensor (thick black line) constrained by an obstacle 13 at point P is shown for each of two stop positions or grid cells. Furthermore, the visible length within the region of interest 15 is shown (thick black line with double arrows), which is constrained in the direction of travel by the visual range or field of view 14 of the sensor, since a forward-pointing sensor cannot or can only detect to a limited extent to the sides or rear of the vehicle 1, and in the opposite direction by the distance to the obstacle 13. This simplification assumes that it is sufficient to see half the lane width up to the end of the region of interest 15. In this case, it is clear that stop positions from grid cells marked white have a significantly higher quality with respect to these parameters, since the line of sight is significantly longer than stop positions starting from grid cells marked with stripes. To calculate the ratio now, for this purpose, the length 15 of the region of interest is first determined, which can be done in a comfortable way (e.g., 2 m / s 2 ) is required for the planned trajectory to include so-called unfavorable scenarios (worst case scenarios) to avoid a collision. Furthermore, the line of sight to the adjacent lane is determined (e.g., covering half of the adjacent lane) and the field of view 14 is possibly simplified (showing the field of view 14 simply as a boundary or line of sight, as exemplified in Figure 7). The overlap of the field of view 14 and the region of interest 15 is then determined and used in the cost function. Grid cells or stopping positions can then be marked accordingly, so that their line of sight lengths are either acceptable or unacceptable. For example, if both vehicles 1, 12 are traveling at 4 m / s 2 , the line of sight can be characterized as unacceptable. Of course, multiple attention regions 15 can be provided here, for example if lane changes to multiple adjacent lanes are possible. Furthermore, the length of the obstacle 13 can also be taken into account in this calculation if it can be determined with high reliability or if it can be estimated based on the object classification.

[0030] Figure 8 shows an undesirable scenario in which an oncoming vehicle 12 appears at the top edge of the attention area 15 when the vehicle 1 starts to move away from a stopped position or to follow an avoidance trajectory. In this scenario, it is important to calculate the required length 15 of the attention area to ensure that a collision with another road user is prevented even with a comfortable braking action (which is significantly slower than an emergency braking action). To determine the length 15 of the attention area, the target or vehicle 12 must be traveling at a certain speed v with a definable tolerance threshold. Target (e.g., the speed allowed by traffic regulations, the speed adopted depending on the traffic situation, etc.) and the vehicle 12 has a conservative or generous set reaction time (t react ) for comfortable braking at speeds (v Target Furthermore, it is important to reduce the vehicle 1's acceleration a Ego ) and then brake comfortably to a speed (v Ego ) (for simplicity, the acceleration values ​​can be assumed to be the same). In particular, the calculation can be performed by the following formula: ROI Length =(t react Ego )2·a Ego +t react Target ·v Target +(v Target ) 2 / (2 a Target )

[0031] 9A and 9B show examples of the transition of the speed of the vehicle 1 or the host vehicle and the oncoming vehicle 12. FIG.

[0032] 10 and 11 show further designs of how the optimal stopping position can be determined. In principle, the stopping position with the best (highest) quality for the maneuver at hand is selected. For this purpose in FIG. 9, a grid map 16 is first prepared, and a stopping position or grid cell is selected as the optimal stopping position (marked as grid cell X), while the vehicle 1 moves towards the optimal stopping position, this stopping position can be saved or stored in a so-called trajectory list. A planned trajectory to the optimal stopping position is then executed so that the vehicle 1 stops at this stopping position X. At this point, the trajectory planning unit of the vehicle 1 can now proceed and determine the requested trajectory or avoidance trajectory, as shown in FIG. 11. The vehicle 1 now follows the avoidance trajectory as soon as the oncoming vehicle 12 has passed and no further oncoming traffic is detected. In particular, the trajectory planning unit can also plan a trajectory at a high repetition rate while driving, thereby controlling the vehicle to return to the starting lane, but if no possible trajectory is found, the vehicle 1 will remain in the avoidance lane until it has passed the obstacle 13 and a trajectory to the starting lane is found.

[0033] Furthermore, a situation in which no oncoming vehicle or oncoming traffic is detected (vehicle 12 is now, for example, located in a lane adjacent to the adjacent lane, or is not present at all) is shown in Figure 12. If sufficient visual range is now provided while approaching optimal stopping position X, vehicle 1 can be controlled in such a way that it already follows the avoidance trajectory a configurable distance (for example, 15 m) to the optimal stopping point or grid cell X. At vehicle level, this maneuver results in first slowing down the speed to optimal stopping position X (for example, speed down to 20 km / h, 30 km / h, 50 km / h, etc.) and then moving the vehicle into the adjacent lane without coming to a complete stop at the optimal stopping position. This design generates a driving style that is particularly similar to real driving behavior, especially in fully automated vehicles.

[0034] FIG. 13 shows the system architecture design of the vehicle assistance system according to the present invention. Sensor data from the surrounding environment sensors is the initial starting point, enabling the detection of the surrounding environment of the vehicle 1 and the execution of surrounding environment interpretation and surrounding environment analysis. The collected and derived information is then summarized into an surrounding environment model 17, which may include an object list listing detected objects (or already classified objects), and sent to a situation interpretation / analysis 18 (situation assessment). Furthermore, a road model or carriageway model may be generated based on the surrounding environment model 17 and transmitted to, for example, a driving planner 19. The driving planner 19 may be configured to determine or calculate the vehicle's driving trajectory. For this purpose, the driving planner 19 includes a maneuver planner 20 and a trajectory planner 21. For this purpose, in particular, the predicted object trajectories from the situation interpretation / analysis 18 may also be used to plan the upcoming maneuver or trajectory. Information about the determined maneuver or trajectory is transferred to a situation manager 22 to determine the current situation and output it to the driver or passengers via a human-machine interface 23 (HMI). Furthermore, the driver or passengers can input or send information or commands to the system via the human-machine interface 23. Initiations or queries can thus be passed to the drive planner 19 via the situation manager 22. The trajectory determined by the drive planner 19 is sent to the motion controller 24, which accesses various vehicle actuators or actuator groups (e.g., steering 3, motor 4, brake 5) to execute the desired trajectory or maneuver. Furthermore, actuator constraints (e.g., physical constraints of the actuators, etc.) or driving dynamics parameters can be sensed by the motion controller 24 and then sent to the drive planner 19 or maneuver planner 20 and / or trajectory planner 21 to be taken into account in the current and / or upcoming plans.

[0035] FIG. 14 shows the operation planner 19 of FIG. 13 in more detail. In this case, a road model, an object list including predicted object movements, etc. are sent as inputs to the operation planner 20 and the trajectory planner 21. The operation planner 20, in this case, includes an operation situation unit 25 (operation situation unit), which determines the current operation situation and requests data from a stop position module 26 (stop pose handler) for this purpose. In this case, the stop position module 26 is used to prepare a stop position matrix (or grid map 16) and request the necessary trajectories for this purpose. The requested trajectories (including trajectories for approaching and departing from stop positions) are then sent to the operation situation unit 25. The operation situation unit 25 then requests the trajectory planner 21 to calculate a trajectory. Furthermore, a trajectory list compiler 27 (trajectory list assembly) is provided, which compiles the currently determined trajectories of the trajectory planner 21 into a trajectory list. Furthermore, the trajectory list compiler 27 determines or calculates the cost of each trajectory, which can be assigned to the trajectory list accordingly. The trajectory list including the determined costs is then sent to a module that evaluates the trajectory list, i.e., the trajectory list evaluator 28 (trajectory list evaluation), so that the trajectories can be evaluated, and at the end of execution, the selected trajectory can be sent to the motion controller 24 for execution. In particular, the trajectory list evaluator 28 is executed on the determined costs of each trajectory, so that the trajectory with the favorable cost ratio is selected and sent to the motion controller 24, which then executes or runs it. The trajectory planner 21 also determines the requested trajectories for the operation planner 20 (or the stop position module 26), and generally one trajectory is determined for each operation.

[0036] For example, the trajectory list may include a set of possible driving trajectories, e.g., one trajectory is provided to stay in the lane, one trajectory is provided to leave the lane, one trajectory is provided to head towards stop position X, and one trajectory is provided to head away from stop position X. The stop position module 26 may also include various trajectories in the trajectory list, e.g., one or more departure trajectories (trajectories) from a central stop position, stop position 1, stop position 2, ... stop position n, etc.

[0037] On the one hand, trajectories can be stored in this list as such and later selected as a driving trajectory, on the other hand, there are also trajectories in the list that have been determined solely for the purpose of determining optimal virtual stopping positions.

[0038] The driving planner 19 (particularly the stop position module 26) can now calculate the visible length of the attention area 15 based on the thick black line with double arrows, as shown in FIG. 7. This calculation can now be determined for each grid cell or stop position. In this case, the object point P closest to the ego-vehicle 1 (center in the x-direction or width of the grid map) and the stop position of the attention area 15 (center in the y-direction or length) is first determined. Furthermore, in this case, the center line of the attention area 15 can be constructed (for example, along the longitudinal center line of the lane). In a next step, a line of sight from the center of the front of the ego-vehicle or vehicle 1 or each sensor (in this example, the sensor located in the center of the front area of ​​the vehicle 1) can be constructed to intersect with point P and extend to the visible length of the attention area 15. The visible length can be calculated or determined in this way. Thereafter, a cost function of the visible length LL, i.e., the cost C1, the length L of the attention area 15, and the like are calculated. ROI can be calculated or determined (e.g., C1 = L L / L ROI). The driving planner 19 (or the stop position module 26) can also determine or calculate a cost value C2 of the distance to the obstacle 13 (e.g., C2 = constant 1 / distance between the vehicle 1 and the obstacle 13). The driving planner 19 can also calculate departure trajectories for all stop positions in the grid map 16 and store them in the stop position module 26 or in a trajectory list in the stop position module 26. Each trajectory may include information on whether it is possible to drive in principle and the cost of the trajectory, taking into account safety, comfort, and route cost. These factors or costs can then be used to determine the quality of each stop position. The safety cost in this case is the total distance to the obstacle 13 or other object, the comfort cost in this case is, among other things, the lateral and / or longitudinal acceleration, the yaw rate, and / or the steering angular velocity, and the route cost in this case can take into account, among other things, how far the vehicle 1 has traveled through the area of ​​interest 15 in a specific time after departing from the stop position. As a result, the driving planner 19 can calculate a cost value for each cell, discarding stops associated with individual cells for which a trajectory cannot or has not been determined. In addition to the cost value calculated to determine the quality of a stop with respect to "driving around obstacles," additional cost values ​​or criteria can be used to discard certain stop locations. For example, costs can be calculated for proximity to construction sites, bicycle paths, or crosswalks, or a stop location can be completely excluded because it is located at a crosswalk or may obstruct a merging lane. Proximity to advantageous areas can also be incorporated into the assessment of stop locations, such as roadside locations required to form emergency lanes.

[0039] FIG. 15 shows the design of a simplified scheme where the individual values ​​of the individual cells can be used for the decision. [Explanation of symbols]

[0040] 1 vehicle 1A vehicle 1B vehicle 2. Control Unit 3 Steering system 4 motors 5. Brakes 6. Front camera 7 Lidar Sensor 8 Radar sensor (long-range radar) 9a~9d Radar sensor (short-range radar) 10a~10d Ultrasonic sensors 11a~11d Surround view camera 12 vehicles 13 Obstacles 14 Field of view 15 Areas of Interest 16 Grid Map 17 Surrounding environment model 18 Situation Interpretation / Situation Analysis (Situation Assessment) 19 Operation Planning Department 20 Operation Planning Department 21 Trajectory Planning Department 22 Situation Management Department 23 Human-Machine Interface (HMI) 24 Motion control section 25 Operational Status Machine 26 Stop position module (stop pause handler) 27 Trajectory List Compiler (Trajectory List Assembly) 28 Trajectory List Assessment (Trajectory List Evaluation)

Claims

1. A method for determining an optimal stopping position for a vehicle (1), comprising the steps of: The vehicle (1) has at least one surrounding environment sensor for detecting surrounding environments, and surrounding environment analysis and / or surrounding environment interpretation are performed to determine traffic situations based on the detected surrounding environments; determining the optimum stopping position in front of the detected obstacle (13) if the vehicle (1) should stop in front of the obstacle (13) due to the traffic situation; a grid map (16) including a plurality of grid cells for determining the optimal stopping position; The grid cells correspond to determined stopping positions and the quality of these stopping positions, The method wherein the optimal stopping position is selected from the stopping positions of the grid map (16) based on the quality of each stopping position.

2. 2. A method according to claim 1, characterized in that a trajectory of the vehicle (1) intended to avoid the obstacle (13) is determined.

3. 3. A method according to claim 1 or 2, characterized in that a trajectory intended to avoid the obstacle (13) can be determined, said trajectory being used to determine an area of ​​interest (15).

4. Method according to any one of claims 1 to 3, characterized in that the field of view (14) of said at least one ambient environment sensor is determined.

5. 5. The method of claim 4, wherein an overlap region of the region of interest (15) and the field of view (14) is determined.

6. Method according to any one of claims 1 to 5, characterized in that a trajectory and / or a field of view (14) and / or a region of interest (15) and / or an overlap region are determined for one or each grid cell.

7. The quality of each said stopping position is determined by the following parameters: the distance to said obstacle (13), and / or the size of said field of view (14), and / or the size of the overlapping area, and / or the size of the line of sight, and / or - the path of the trajectory that reaches said stopping position, and / or - the path of the trajectory leaving said stopping position, and / or the comfort of the trajectory to reach said stopping point, and / or the comfort of the trajectory leaving said stopping position, and / or the safety of the trajectory to reach said stopping point, and / or the safety of the trajectory leaving said stopping position, and / or - the drivability of the trajectory leaving said stop position, and / or the distance from the disadvantageous area to said stopping position, and / or - the distance of said stopping position from the advantageous area The method according to any one of claims 1 to 6, characterized in that it is defined based on

8. 8. The method according to claim 1, wherein the cost of at least one parameter is determined based on a cost function and / or the cost of at least one trajectory is determined and then used to determine the quality.

9. An assistance system for a vehicle (1), comprising: a control unit (2) for controlling the vehicle (1); at least one sensor for sensing the surrounding environment; Including, an environment analysis and / or environment interpretation for determining the traffic situation is performed based on the detected environment; If the vehicle (1) has to stop in front of the obstacle (13) due to the traffic situation, an optimal stopping position of the vehicle (1) in front of the detected obstacle (13) is determined by a method according to any one of claims 1 to 8. Support system.

10. A vehicle (1) comprising an assistance system according to claim 9.

Images