Method for guiding a vehicle to a limited-visibility traffic node

Abstract

The invention discloses a method for slow driving a vehicle to a target position, in particular to a traffic node, by means of a controller, wherein sensor data of a vehicle sensor arrangement and / or of an infrastructure-side sensor arrangement are received while the vehicle is parked at a stop position, a risk / utility calculation is performed by means of the received sensor data for driving beyond the stop position and for slow driving to a target position arranged behind the stop position in the driving direction, the vehicle is driven beyond the stop position when it is safe for the vehicle to drive forward and / or beyond the stop position, and the stepwise approach of the vehicle from the stop position to the target position is started when the result of the risk / utility calculation becomes greater compared to the result of the risk / utility calculation at the stop position. Furthermore, a controller, a computer program and a computer-readable storage medium are disclosed.

Description

Technical Field

[0001] This invention relates to a method for controlling a vehicle to slowly move towards a target location, particularly towards a traffic node, using a controller. Furthermore, the invention also relates to a controller, a computer program, and a machine-readable storage medium. Background Technology

[0002] In urban traffic, autonomous vehicles should also be able to turn or merge into traffic gaps when they do not have the right-of-way. Driving without the right-of-way is also considered an unsafe driving maneuver under this relationship.

[0003] In most traffic conditions, autonomous vehicles need to stop at the stop line or designated stopping position before initiating unsafe driving maneuvers.

[0004] This approach may be disadvantageous because the detection range of vehicle sensors at stop lines is often limited by obstacles. Furthermore, traffic conditions with increased flow can further impede the detection range of vehicle sensors and make future driving maneuvers such as merging into traffic gaps more difficult. Summary of the Invention

[0005] The objective of this invention is to propose a method for maximizing the detection range of vehicle sensors and for improving traffic safety.

[0006] This task is solved by means of the corresponding subject matter of the independent claims. The advantageous embodiments of the invention are the subject matter of the dependent claims.

[0007] According to one aspect of the invention, a method is provided for slowing a vehicle to a target location, particularly to a traffic node, via a controller. The controller may be responsible for vehicle behavior planning or may be designed as a behavior planning module.

[0008] Traffic nodes can be, for example, intersections, approach roads, roundabouts, etc. Traffic nodes can be seen to a limited extent from the parking positions of vehicles before they pass through them.

[0009] The controller can start or stop the automatic operation of vehicles, or their slow movement or approach to the stop line or stop position. In particular, the controller ensures compliant control of vehicles capable of automatic operation.

[0010] In one step, when the vehicle stops at the designated stop position, sensor data from the vehicle's sensors and / or the infrastructure's sensors are received. Here, it can be checked whether the detection range of the vehicle's sensors is limited or obstructed by objects or obstacles. This can be determined by identifying the objects using the received sensor data. The stop position can preferably be at the stop line, which is detected by evaluating the sensor data before the vehicle crosses the traffic node.

[0011] Using the received sensor data, risk / utility calculations are performed on the target positions that are beyond the parking position and the target positions that are arranged after the parking position along the driving direction.

[0012] In risk / utility calculation, risk and utility are determined based on the distance from the stopping position and then compared. This comparison of risk and utility can be done by subtraction or algorithmically. The result of the risk / utility calculation can be obtained by comparing risk and utility. If the result of the risk / utility calculation is shown in a spacing-distance table, then the result preferably has a maximum value, which indicates the particularly optimal ratio of utility to risk.

[0013] Preferably, the target location can be determined at this distance from the stop location, where the value of the spacing-distance table or the result of the risk / utility calculation is greater than that at the stop location.

[0014] Alternatively or additionally, the target position can be determined at this distance from the stop position, where the maximum value of the distance-distance table is located.

[0015] For example, the largest possible size of the detection range can be considered as utility and the collision probability as risk. Risk can here be considered as a safety factor in the risk / utility calculation, in the opposite form. The collision probability and the size or dimension of the available detection range can be abstracted and optionally weighted to achieve hedging or comparison.

[0016] When the size of the detection range is as close as possible to the theoretically maximum possible size of the surrounding environment sensing device and the probability of collision is negligible, the result of the risk / utility calculation can be, for example, the maximum value.

[0017] When calculating the risk / utility for a target location, one can take into account how far the vehicle has entered the lane of cross traffic and thus weigh the potential collision or accident risk relative to the increased size of the detection range.

[0018] Preferably, the vehicle only exceeds the stop position if it is safe for the vehicle to proceed forward and / or to exceed the stop position. This prevents the slow approach into a traffic node from becoming a safety risk for the vehicle. Checking whether exceeding the stop position is safe for the vehicle can be done, for example, by evaluating measurements involving nearby traffic participants and determining the collision risk with said or nearby traffic participants. Exceeding the stop position is considered safe, for example, if the determined collision risk is below a predetermined threshold.

[0019] Subsequently, when the risk / utility calculation result improves compared to the risk / utility calculation result at the stop position, the vehicle begins to gradually approach at least one target position from the stop position. Therefore, the vehicle can only cross the stop position or stop line when the size of the vehicle sensor's detection range is improved or expanded compared to the size of the vehicle sensor's detection range at the stop line. To make this decision, the controller can simulate the area or volume of the vehicle sensor's detection range for different target positions spaced from the stop line. A comparison of the corresponding simulated visibility can then be performed. Identifying approaching traffic participants as early as possible can be considered as a criterion for this comparison. During the simulation, the size of the vehicle sensor's detection range at different target positions can be determined and used for the risk / utility calculation. The collision risk calculation can here be compared with the utility of the increased size of the detection range at the target position to obtain the result of the risk / utility calculation.

[0020] The result of a risk / utility calculation can, for example, have any value within the range of 0 to 1 or 0% to 100%. A particularly positive result of a risk / utility calculation can have a value of 1 or 100%, and a particularly negative result has a value of 0 or 0%. Here, under a positive result of a risk / utility calculation, the advantage or utility dominates. Under a negative result of a risk / utility calculation, the risk, especially the potential risk of collision, dominates.

[0021] In risk / utility calculations, one can particularly weigh whether the advantage of a larger detection range for vehicle sensors justifies a higher risk of accidents when crossing lanes of cross traffic.

[0022] This method allows autonomous vehicles to cautiously approach poorly visible or obstructed traffic junctions, preventing unsafe driving maneuvers. In particular, it enables the early detection of traffic participants, providing a sufficient time window for maneuver. By cautiously approaching lanes intersecting with the vehicle, it allows for faster merging into traffic gaps. Furthermore, this method can prevent prolonged congestion at intersections.

[0023] The detection range of vehicle sensors is designed in circular, elliptical, rectangular, triangular, or polygonal shapes around the unrestricted area of ​​the vehicle. The size of the detection range or scanning range of a vehicle sensor is specifically determined by the maximum operating radius and scanning angle of the corresponding vehicle sensor. Depending on the surrounding environment, the detection range may be restricted or obstructed, thus reducing its size.

[0024] The method can preferably be implemented as a general and real-time solution that takes into account the limitations of the detection range of static and / or dynamic obstacle and vehicle sensors.

[0025] The method can be used, for example, in various unsafe driving maneuvers. Such unsafe driving maneuvers may include making a right or left turn during a red light, merging from an auxiliary lane into or into a main lane, unsafely crossing a road intersection, or driving through an intersection without traffic signs.

[0026] Furthermore, the automated vehicle can utilize the method to implement human actions by cautiously approaching traffic nodes, which optimizes the visibility of the vehicle's sensors and enables dynamic vehicle operation while adhering to safety requirements.

[0027] According to another aspect of the invention, a controller is provided, wherein the controller is configured to implement the method. The controller may be, for example, a vehicle-side controller, an external vehicle controller, or an external server unit, such as a cloud system.

[0028] Furthermore, according to one aspect of the invention, a computer program is also provided, comprising instructions that, when executed by a computer or controller, cause the computer to perform the method according to the invention. According to another aspect of the invention, a machine-readable storage medium is also provided, on which the computer program according to the invention is stored.

[0029] The vehicles can operate assistedly, partially automatically, highly automatically, and / or fully automatically, or unmanned, according to BASt standards. Vehicles can be designed as land vehicles, such as robots, drones, rail vehicles, autonomous taxis, industrial robots, commercial vehicles, and buses; as aerial vehicles; or as water vehicles.

[0030] In one embodiment, risk and utility are determined based on the distance between the stop positions. The distance is then calculated to find the maximum value of the offsetting utility and risk, in the form of a risk / utility calculation. In another step, a target position is determined at a distance from the stop position, at which the risk / utility calculation results constitute the maximum value. The controller can then generate corresponding control commands to move towards the determined target position. Thus, the distance can be continuously or in defined steps expanded from the stop position, and the corresponding risk / utility calculation results can be calculated for each distance. This can be accomplished by compiling a result-distance table, which plots the distance-related results of the risk / utility calculation. If a maximum value is detected in the result-distance table, further expansion of the distance can be interrupted, and the distance at the maximum value is used to set the target position from the stop position.

[0031] Calculate the results of risk / utility calculations in relation to the situation.

[0032] In particular, risk / utility calculations can be performed in real time for different target locations by observing vehicle movement or by simulating the detection range of vehicle sensors at the corresponding target locations. Risk / utility calculations can optimally provide an assessment of the usability of vehicle sensors. For example, the risk / utility calculation of a vehicle sensor severely obstructed by objects is worse than that of an unobstructed vehicle sensor. The maximum size of the vehicle sensor's detection range and the likelihood of detecting distant traffic participants play a particularly decisive role in the risk / utility calculation.

[0033] The availability of a vehicle sensor can be determined, for example, by comparing the current or simulated size of the detection range of the vehicle sensor with the theoretically unobstructed or maximum possible size of the detection range.

[0034] The target location can be established continuously or in defined or discrete steps. Each step can be associated with a spacing of, for example, a few centimeters or meters between intersecting lanes and stop lines.

[0035] The optimal detection range of a vehicle sensor across one or more lanes may be important for calculating risk / utility. Lanes that cross vehicles, for example, can be used as lanes important for risk / utility calculations in order to detect traffic participants as early as possible.

[0036] In another implementation, upon reaching the target location and / or when the risk / utility calculation reaches its maximum result, control commands are generated for stopping the vehicle or for performing an entry operation into a road intersecting with the vehicle's lane. The entry operation can here be an unsafe driving maneuver. Reaching the maximum result of the risk / utility calculation allows for particularly rapid merging or entry into a new lane and / or enables a particularly large size of the detection range of the vehicle sensors related to the planned lane.

[0037] According to another embodiment, when the vehicle exceeds the stop position, a risk / utility calculation is continuously performed by the vehicle. Preferably, different detection ranges of the vehicle's sensors at possible target locations are determined to identify collision risks, wherein the target position is calculated based on the maximum size of the detection range of the vehicle's sensors, and the vehicle proceeds towards this target position. A control command is generated for this controller position, and the vehicle proceeds towards the fixed target position by means of the control command.

[0038] It is preferable to drive towards one or more fixed target locations at a low speed, such as walking speed. By setting the target locations, the possible detection range of the vehicle's sensors can be evaluated at different locations. Furthermore, the risks and benefits for the vehicle can be determined at any target location. Preferably, the vehicle can be driven towards a target location where the risks are minimized and the benefits are maximized.

[0039] In another implementation, slow movement towards the target location begins when the detection range of the vehicle sensor is expected to increase and / or entry into the lane intersecting with the vehicle's lane is expected to be simplified. This measure, in addition to expanding the detection range of the vehicle sensor, also ensures particularly simple and rapid entry into the lane, serving as a decisive criterion for risk / utility calculations. The expansion of the detection range can be achieved, for example, by increasing the area and / or volume of the detection range. Specifically, the controller can introduce control commands to initiate unsafe driving maneuvers when traffic conditions at the previously reached target location permit such maneuvers.

[0040] In another embodiment, risk / utility calculations are performed using sensor data from the vehicle's own sensors, infrastructure-side sensors, and / or sensors from traffic participants surrounding the vehicle. This allows the controller to receive and evaluate a large amount of sensor data to determine the optimal target location and guide the vehicle to or into a traffic node with minimal risk. Sensor data can be transmitted, in particular, via one or more communication connections, such as Car-2-X or Car-2-Car communication connections. By using sensor data from other traffic participants and infrastructure units such as traffic lights, the controller can obtain a panoramic view of the traffic situation that is less obstructed by objects or obstacles.

[0041] In another implementation, the presence of people around the vehicle is checked, minimizing harm to people or objects in the vicinity before the vehicle slows to the target location. For example, if a person or cyclist wants to cross the vehicle's lane, the controller can prevent them from crossing the stop position. Furthermore, the target location can be prevented from being set on a sidewalk or zebra crossing.

[0042] In another embodiment, information about the vehicle's characteristics and traffic rules are stored in the controller and taken into account during the risk / utility calculation. This measure ensures compliant vehicle direction even during unsafe driving maneuvers. For example, the vehicle can stop at a stop sign before proceeding to a further destination. Furthermore, it avoids interfering with lane-jumping or higher-priority traffic participants.

[0043] It is preferable to set the speed of a vehicle heading toward at least one target location so low that traffic participants crossing the vehicle will not be startled by it and thus react by initiating braking. For this purpose, the vehicle may proceed toward the target location particularly slowly or at a decreasing speed. Attached Figure Description

[0044] The preferred embodiments of the invention will now be explained in more detail with the aid of strongly simplified schematic diagrams. In the diagrams:

[0045] Figure 1-3 This is a top view illustrating a first traffic situation used to explain a method of slowly advancing a vehicle to a target location according to one embodiment;

[0046] Figure 4-5 This is a top view used to illustrate a second traffic situation in accordance with one embodiment of a method for slowly advancing a vehicle to a target location;

[0047] Figure 6 This is a schematic flowchart used to check if the stop position has been exceeded. Detailed Implementation

[0048] Figures 1 to 3 A top view is shown illustrating a first traffic situation 1 for explaining a method of causing vehicle 2 to slowly advance to a target position Z according to one embodiment. The target position Z is determined by the controller 4 on the vehicle side, having the maximum risk / utility calculation or the maximum result of said risk / utility calculation. Vehicle 2 may travel towards the target position Z in the direction of travel F. Under the maximum risk / utility calculation, the utility or size of the detection range 18 is maximized, and the risk to vehicle 2, for example, due to collisions with other road users, is minimized.

[0049] To calculate the risk / utility, controller 4 can receive sensor data from vehicle sensing device 6. Additionally, controller 4 can receive sensor data from traffic participants 8 and / or infrastructure units 10. For this purpose, a communication connection 12 can be established to receive sensor data calculated outside the vehicle. The detection range 18, or visibility, of vehicle sensing device 6 is schematically shown.

[0050] In the illustrated embodiment, the vehicle sensor 6 is obstructed by a stationary obstacle 14. This may prevent a traffic participant 8 crossing the lane 16 of the vehicle 2 from being seen by the vehicle sensor 6. This traffic participant 8 may alternatively or additionally be a dynamic obstacle.

[0051] Figure 1 Vehicle 2 is shown stopped at position S. Position S corresponds to the stop line. At position S, the visibility 18 of vehicle sensor 6 of vehicle 2 is obstructed by obstacle 14. In particular, lane 16 intersecting with vehicle 2 is not visible.

[0052] For this reason, the target position Z is determined by controller 4, and the distance A between the target position and the stop position S is [missing information]. This step is [missing information]. Figure 2 and Figure 3 This is explained in the text. Here, the target position Z is continuously shifted forward and the spacing A increases until the visibility 18 of the vehicle sensor 6 over lane 16 is at its maximum. Figure 3 In this way, vehicle 2 moves forward, allowing traffic participants 8 to be scanned or "seen" in advance by vehicle sensor device 6.

[0053] During this process, the impact of other traffic participants 8 on the visibility 18 of the vehicle sensor 6 can be continuously examined. The predicted positions of traffic participants 8 and obstacles 14 relative to the vehicle 2 can be simulated by the controller 4. This allows for the determination of the optimal target position Z.

[0054] exist Figure 3In this way, vehicle 2 arrives at the target position Z that has been pushed forward, allowing vehicle 2 to turn into lane 16 clearly and without risk.

[0055] Figure 4 and Figure 5 A top view is shown illustrating a second traffic situation 1 for illustrating a method of slowing a vehicle 2 to a target position Z according to one embodiment.

[0056] As illustrated in Figure 5, the new target position Z of vehicle 2 may pose a collision risk because vehicle 2 has entered lane 16 of traffic participant 8. Once such a collision risk with traffic participant 8 crossing vehicle 2 exists, the corresponding target position Z can be abandoned.

[0057] The controller 4 can take into account the vehicle size, the lane width B, and the angle W when calculating the target position Z.

[0058] Figure 6 A schematic flowchart for checking if the stop position S has been exceeded is shown. The check can be performed via controller 4. In particular, it checks whether a slow approach to the target position Z is permissible.

[0059] When vehicle 2 is parked at the stop position S, vehicle 2 can be slowly moving forward or so-called "slow advance" 20 by the control command of controller 4.

[0060] In traffic condition 1, where unsafe intersections or turns are not permitted, vehicle 2 remains at a stopped position S, where the target position Z for slow progress is not calculated.

[0061] Here, the controller can start from the vehicle 2's stopping position S, 22, or from the target position Z, or from a slow movement 20, and put the vehicle 2 into a driving mode 24. The vehicle turns into lane 16 through this driving mode and can follow this lane. This can be done, for example, when sufficient clearance is found between traffic participants 8.

Claims

1. Method for creeping a vehicle (2) to a target position (Z) by means of a controller (4), wherein - sensor data of a vehicle sensor device (6) and / or of an infrastructure-side sensor device (10) are received while the vehicle (2) is parked at a stop position (S), - a risk / effort calculation is performed by means of the received sensor data for passing the stop position (S) and for creeping to a target position (Z) arranged behind the stop position (S) in the driving direction (F), - the stop position (S) is passed when the vehicle (2) drives forward and / or passing the stop position (S) is safe for the vehicle (2), - a stepwise approach of the vehicle (2) from the stop position (S) to the target position (Z) is started when the result of the risk / effort calculation becomes greater compared to the result of the risk / effort calculation at the stop position (S), 2. The method of claim 1, wherein, - the risk and the effort are determined depending on a distance (A) to the stop position (S), wherein the distance (A) is found at which there is a maximum of a hedge of the effort and the risk in the form of the result of the risk / effort calculation, and wherein the target position (Z) is determined at a distance (A) to the stop position (S) at which the result of the risk / effort calculation constitutes the maximum.

3. The method according to claim 1 or 2, wherein, The method is used for creeping a vehicle (2) to a target position (Z) in front of a traffic node by means of a controller (4).

4. The method according to any one of claims 1 to 3, wherein, Upon reaching the target position (Z) and / or upon the greatest result of the risk / effort calculation, a control command is generated for parking the vehicle (2) or for performing a pull-in maneuver (24) onto a road (16) crossing the lane of the vehicle (2).

5. The method according to any one of claims 1 to 4, wherein, Upon passing the stop position (S), a risk / effort calculation is continuously performed by the vehicle (2), wherein different sizes of a detection range (18) of the vehicle sensor device (6) at possible target positions (Z) are found to identify a collision risk and based thereon the target position (Z) is calculated on the detection range (18) of the vehicle sensor device (6) having the greatest size.

6. The method according to any one of claims 1 to 5, wherein, Creeping to the target position (Z) is started when the size of the detection range (18) of the vehicle sensor device (6) is expected to become greater and / or a pull-in onto a lane (16) crossing the lane of the vehicle is expected to be simplified.

7. The method according to any one of claims 1 to 6, wherein, The risk / effort calculation is performed using sensor data of the vehicle sensor device (6) of the vehicle (2), of the infrastructure-side sensor device (10) and / or of a vehicle sensor device of a traffic participant (8) from the surroundings of the vehicle (2).

8. The method according to any one of claims 1 to 7, wherein, It is checked whether there are persons in the surroundings of the vehicle (2), wherein a hazard to persons or objects in the surroundings of the vehicle (2) is minimized before the vehicle (2) creeps to the target position (Z).

9. Controller (4), wherein Information about properties of the vehicle (2) and traffic rules are stored in the controller (4) and taken into account when the risk / effort calculation. The controller (4) is provided for carrying out the method according to any one of claims 1 to 8.

10. Computer program product comprising a computer program comprising instructions which, when the computer program is executed by a computer or controller (4), cause the computer to carry out the method according to any one of claims 1 to 8.

11. Machine-readable storage medium on which the computer program of the computer program product according to claim 10 is stored.

Images