CN-122015895-A - Unstructured road-oriented obstacle avoidance path planning method, device and storage medium

CN122015895ACN 122015895 ACN122015895 ACN 122015895ACN-122015895-A

Abstract

The invention provides an unstructured road-oriented obstacle avoidance path planning method, device and storage medium, and relates to the field of automatic driving. The method comprises the steps of acquiring information of an obstacle if the obstacle is encountered in the process of running of a vehicle according to a reference path, determining a plurality of candidate obstacle avoidance paths for the vehicle to avoid the obstacle according to the reference path, the information of the obstacle, the size information and the pose information of the vehicle, for the obstacle which does not meet the crossing condition of the vehicle, the plurality of candidate obstacle avoidance paths comprise candidate obstacle avoidance paths in a detour mode, for the obstacle which meets the crossing condition of the vehicle, the plurality of candidate obstacle avoidance paths comprise candidate obstacle avoidance paths in the crossing mode and the detour mode, selecting a target obstacle avoidance path from the plurality of candidate obstacle avoidance paths according to the passing cost of each candidate obstacle avoidance path, and updating the reference path based on the target obstacle avoidance path. The invention can improve the driving safety and the passing efficiency of the automatic driving vehicle in the complex unstructured road.

Inventors

Guo Congshuai
LI SHAOHUA
ZHANG ZHE
FENG GUIZHEN
WANG XUEWEI
SUN ZHIQI

Assignees

石家庄铁道大学

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. An unstructured road-oriented obstacle avoidance path planning method is characterized by comprising the following steps: in the process that the vehicle runs according to a reference path provided by a navigation system, if an obstacle is encountered, information of the obstacle is collected; Determining a plurality of candidate obstacle avoidance paths for the vehicle to avoid the obstacle according to the reference path, the information of the obstacle, the size information and the pose information of the vehicle, wherein the plurality of candidate obstacle avoidance paths comprise candidate obstacle avoidance paths in a detour mode for the obstacle which does not meet the crossing condition of the vehicle; And selecting a target obstacle avoidance path from the plurality of candidate obstacle avoidance paths according to the passing cost of each candidate obstacle avoidance path, and updating the reference path based on the target obstacle avoidance path.
2. The method for planning an obstacle avoidance path for an unstructured roadway according to claim 1, wherein the determining a plurality of candidate obstacle avoidance paths for the vehicle to avoid the obstacle according to the reference path, the information of the obstacle, the size information and the pose information of the vehicle comprises: determining a plurality of left-hand winding key points and a plurality of right-hand winding key points of the obstacle according to the information of the reference path and the obstacle; Generating a plurality of candidate obstacle avoidance paths in a detour mode according to the left-hand key points and/or the right-hand key points and the pose information of the vehicle; acquiring a crossing condition of the vehicle according to the size information and the pose information of the vehicle; Judging whether the obstacle meets the crossing condition of the vehicle according to the information of the obstacle; If the obstacle meets the crossing condition of the vehicle, determining a plurality of obstacle crossing key points of the obstacle according to the information of the reference path and the obstacle, and generating a plurality of candidate obstacle avoidance paths in a crossing mode according to the obstacle crossing key points and the pose information of the vehicle.
3. The unstructured road-oriented obstacle avoidance path planning method of claim 2 wherein the information of the obstacle comprises a center point and a direction angle, wherein the plurality of left-hand-wound keypoints comprises a first left-hand-wound keypoint, a second left-hand-wound keypoint, and a third left-hand-wound keypoint, and wherein the plurality of right-hand-wound keypoints comprises a first right-hand-wound keypoint, a second right-hand-wound keypoint, and a third right-hand-wound keypoint; the determining a plurality of left-winding key points and a plurality of right-winding key points of the obstacle according to the information of the reference path and the obstacle comprises the following steps: Determining the safety avoidance length and the safety avoidance width of the obstacle according to the tangential angle of the reference path, the preset transverse safety distance and the information of the obstacle; Establishing a parallelogram which takes the center point of the obstacle as the center and takes the safety avoidance length and the safety avoidance width of the obstacle as the side length as a first safety avoidance bounding box, wherein the included angle of the first safety avoidance bounding box is obtained according to the direction angle of the obstacle; Determining the first left-hand key point, the second left-hand key point, the first right-hand key point and the second right-hand key point according to the vertex of the first safety avoidance bounding box; Determining a lateral detour distance of the obstacle according to the tangential angle of the reference path, the direction angle of the obstacle and the safety detour distance of the obstacle, and determining a second safety detour bounding box according to the first safety detour bounding box and the lateral detour distance of the obstacle, wherein the included angle of the second safety detour bounding box is obtained according to the tangential angle of the reference path; and determining the third left-winding key point and the third right-winding key point according to the vertex of the second safety avoidance bounding box.
4. The unstructured road-oriented obstacle avoidance path planning method of claim 3 wherein the plurality of obstacle-crossing keypoints comprises a first obstacle-crossing keypoint, a second obstacle-crossing keypoint, a third obstacle-crossing keypoint, and a fourth obstacle-crossing keypoint; The determining a plurality of obstacle crossing key points of the obstacle according to the information of the reference path and the obstacle comprises the following steps: Determining the first obstacle crossing key point and the second obstacle crossing key point according to the central point of the obstacle, the safety avoidance length and the direction angle; determining the third obstacle crossing key point and the fourth obstacle crossing key point according to the center point of the obstacle, the safety avoidance length and the tangential angle of the reference path; the connecting line of the first obstacle crossing key point and the second obstacle crossing key point is parallel to the direction of the obstacle, and the connecting line of the third obstacle crossing key point and the fourth obstacle crossing key point is parallel to the tangential direction of the reference path.
5. The method for planning an obstacle avoidance path for an unstructured road according to claim 4, wherein generating a plurality of candidate obstacle avoidance paths in a detour mode according to the left-hand and/or right-hand turn keypoints and pose information of the vehicle comprises: Selecting a starting point and an ending point of a first interval path from the left-hand key points or the right-hand key points, and connecting the starting point and the ending point of the first interval path by line segments to obtain the first interval path, wherein the starting point and the ending point of the first interval path belong to the left-hand key points or the right-hand key points and are respectively positioned on a vehicle driving side and a vehicle driving away side of the obstacle; Generating a second interval path from a preset obstacle avoidance starting point to the starting point of the first interval path and a third interval path from the end point of the first interval path to a preset obstacle avoidance end point according to the pose information of the vehicle; Connecting the first interval path, the second interval path and the third interval path to obtain a candidate obstacle avoidance path in a detour mode; Repeatedly executing the operations to generate a plurality of candidate obstacle avoidance paths in a detour mode; The generating a plurality of candidate obstacle avoidance paths in a crossing mode according to the obstacle crossing key points and the pose information of the vehicle comprises the following steps: Selecting a starting point and an ending point of a first interval path from the obstacle crossing key points, and connecting the starting point and the ending point of the first interval path by line segments to obtain the first interval path, wherein the starting point and the ending point of the first interval path are respectively positioned on a vehicle driving side and a vehicle driving away side of the obstacle; Generating a second interval path from a preset obstacle avoidance starting point to the starting point of the first interval path and a third interval path from the end point of the first interval path to a preset obstacle avoidance end point according to the pose information of the vehicle; Connecting the first interval path, the second interval path and the third interval path to obtain a candidate obstacle avoidance path in a crossing mode; And repeatedly executing the operations to generate a plurality of candidate obstacle avoidance paths in the crossing mode.
6. The method for planning an obstacle avoidance path for an unstructured road according to claim 1, wherein selecting a target obstacle avoidance path from the plurality of candidate obstacle avoidance paths according to the traffic cost of each candidate obstacle avoidance path comprises: Constructing a passing cost function of the candidate obstacle avoidance path, wherein the passing cost function comprises a driving safety cost based on an obstacle risk potential field, a lateral stability cost based on the curvature of the candidate obstacle avoidance path and a lateral offset cost based on the lateral distance difference between the candidate obstacle avoidance path and the reference path; and evaluating the plurality of candidate obstacle avoidance paths based on the traffic cost function, and selecting the candidate obstacle avoidance path with the minimum traffic cost as the target obstacle avoidance path.
7. The unstructured road-oriented obstacle avoidance path planning method of claim 6 wherein the risk value of the obstacle risk potential field increases as the distance between the vehicle and the obstacle decreases when the vehicle is within the range of influence of the obstacle risk potential field, and wherein the risk value of the obstacle risk potential field is zero when the vehicle is outside the range of influence of the obstacle risk potential field; the influence range of the obstacle risk potential field is determined according to the size information of the vehicle, the braking safety distance and the information of the obstacle.
8. The unstructured road-oriented obstacle avoidance path planning method of any of claims 1-7, wherein after updating the reference path, the method further comprises: Constructing an objective function related to an updated reference path, solving the objective function based on a nonlinear vehicle kinematic prediction model and vehicle running constraint to obtain a vehicle optimal control input in a prediction time domain of the nonlinear vehicle kinematic prediction model, and performing discretization modeling on a prediction time domain of the nonlinear vehicle kinematic prediction model by using at least two different sampling step sizes in the process of solving the objective function, wherein the sampling step size close to the current moment is smaller than the sampling step size far from the current moment; And generating an optimal obstacle avoidance path corresponding to the optimal control input of the vehicle based on the nonlinear vehicle kinematics prediction model.
9. An electronic device comprising a memory storing a computer program and a processor implementing the method of any one of claims 1 to 8 when the computer program is executed by the processor.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 8.

Description

Unstructured road-oriented obstacle avoidance path planning method, device and storage medium Technical Field The invention relates to the field of automatic driving, in particular to an obstacle avoidance path planning method, equipment and storage medium for unstructured roads. Background The automatic driving off-road vehicle equipment is widely used in a plurality of fields of industry and agriculture, and has wide development prospect and remarkable economic value. In unstructured road environments, such as open country, off road or unpaved areas, roads lack clear lane lines, the types of obstacles are diverse and randomly distributed, and a significant challenge is presented to the path planning of an autonomous vehicle. In such a scene, path planning needs to consider both driving efficiency and track stability under the premise of ensuring safety so as to adapt to complex terrain conditions. The existing automatic driving vehicle path planning method mostly adopts a fixed rule obstacle avoidance strategy, for example, a bypass path is generated based on a manual potential field or a sampling search mode. The methods generally preset unified obstacle avoidance behaviors, and are difficult to conduct differentiated treatment according to obstacle characteristics and vehicle states. Because the prior art lacks of fine distinction of obstacle avoidance strategies, an automatic driving vehicle is difficult to adaptively select an optimal passing strategy in complex terrains, and high-efficiency and safe operation of the vehicle in an unstructured environment is restricted. Therefore, there is a need for an obstacle avoidance path planning method capable of adaptively determining an obstacle avoidance strategy and generating a corresponding obstacle avoidance path, so as to improve the driving safety and the traffic efficiency of an automatic driving vehicle on a complex unstructured road. Disclosure of Invention In view of the above, the embodiments of the present invention provide a method, an apparatus, and a storage medium for planning an obstacle avoidance path for an unstructured road, so as to improve the driving safety and the traffic efficiency of an automatic driving vehicle in a complex unstructured road. In a first aspect, an embodiment of the present invention provides a method for planning an obstacle avoidance path for an unstructured road, including: in the process that the vehicle runs according to a reference path provided by a navigation system, if an obstacle is encountered, information of the obstacle is collected; Determining a plurality of candidate obstacle avoidance paths for the vehicle to avoid the obstacle according to the reference path, the information of the obstacle, the size information and the pose information of the vehicle, wherein the plurality of candidate obstacle avoidance paths comprise candidate obstacle avoidance paths in a detour mode for the obstacle which does not meet the crossing condition of the vehicle; And selecting a target obstacle avoidance path from the plurality of candidate obstacle avoidance paths according to the passing cost of each candidate obstacle avoidance path, and updating the reference path based on the target obstacle avoidance path. In one possible implementation manner, the determining a plurality of candidate obstacle avoidance paths for the vehicle to avoid the obstacle according to the reference path, the information of the obstacle, the size information and the pose information of the vehicle includes: determining a plurality of left-hand winding key points and a plurality of right-hand winding key points of the obstacle according to the information of the reference path and the obstacle; Generating a plurality of candidate obstacle avoidance paths in a detour mode according to the left-hand key points and/or the right-hand key points and the pose information of the vehicle; acquiring a crossing condition of the vehicle according to the size information and the pose information of the vehicle; Judging whether the obstacle meets the crossing condition of the vehicle according to the information of the obstacle; If the obstacle meets the crossing condition of the vehicle, determining a plurality of obstacle crossing key points of the obstacle according to the information of the reference path and the obstacle, and generating a plurality of candidate obstacle avoidance paths in a crossing mode according to the obstacle crossing key points and the pose information of the vehicle. In one possible implementation, the information of the obstacle comprises a center point and a direction angle, the plurality of left-hand-wound keypoints comprises a first left-hand-wound keypoint, a second left-hand-wound keypoint and a third left-hand-wound keypoint, and the plurality of right-hand-wound keypoints comprises a first right-hand-wound keypoint, a second right-hand-wound keypoint and a third right-hand-wound keypoint; the determining a pluralit