CN-121612335-B - Global path planning method based on potential energy guiding rapid searching random tree algorithm

CN121612335BCN 121612335 BCN121612335 BCN 121612335BCN-121612335-B

Abstract

Global path planning method based on potential energy guiding rapid searching random tree algorithm. Belongs to the technical field of intelligent automobile path planning. The method solves the technical problem that the traditional rapid search random tree algorithm can cause the planned track to be closely attached to the obstacle for running. Firstly, defining the positions of obstacles and the positions of starting points and ending points through a two-dimensional space map planned by a vehicle, calculating potential energy values of any point in the map, superposing the potential energy values of all the obstacles to obtain potential energy values of any point, obtaining potential energy values of all any point in the two-dimensional space map according to the same method, normalizing the potential energy values of each point of the whole map to obtain a potential energy map matrix, introducing the potential energy map matrix into a fast search random tree algorithm, setting potential energy threshold values, screening randomly generated points, introducing the potential energy values into a cost function of the fast search random tree algorithm after screening, and re-wiring and optimizing to plan a vehicle path.

Inventors

HAN JIAYI
YAN ZIXUAN
SONG DONGJIAN
YANG XIN
XU WENHAO
JIANG YUNCHUAN

Assignees

吉林大学

Dates

Publication Date: 20260508
Application Date: 20260203

Claims (7)

1. The global path planning method based on potential energy guiding rapid searching random tree algorithm is characterized in that the method firstly determines the position of an obstacle and the position of a starting point and an end point according to a two-dimensional space map planned by a vehicle, and aims at any point in the map Calculating potential energy values of the obstacle relative to each obstacle, and superposing the potential energy values of the obstacle relative to all the obstacles to obtain any point According to the same method, obtaining potential energy values of all arbitrary points in a two-dimensional space map, normalizing the potential energy values of each point of the whole map, wherein the normalized numerical value is between 0 and 1, so as to obtain a potential energy map matrix, introducing the potential energy map matrix into a fast search random tree algorithm, setting a potential energy threshold value in an initial stage of path planning, screening random generation points, introducing the potential energy values into a cost function of the fast search random tree algorithm after screening, and re-wiring and optimizing, thereby carrying out vehicle path planning, and the method comprises the following specific steps: s1, initializing a two-dimensional space map of vehicle planning, setting a starting point X_start and a target point X_ goal, loading barrier information, simultaneously importing an environment potential energy map matrix, and setting a maximum iteration number K, an expansion step length, a neighborhood search radius and a potential energy safety threshold value ; S2, constraint sampling based on potential energy guidance, namely randomly sampling in a map space to generate a candidate point X_rand, and reading a potential energy value corresponding to the point in a potential energy map matrix Judging Whether or not it is less than a set potential energy threshold : If it is Judging that the point is in a high risk area, directly discarding the sampling point and re-sampling; If it is Reserving the point as an effective sampling point and entering the next step; S3, node extension and collision detection, namely traversing all nodes in a current random tree, searching a tree node X_nearest which is closest to X_rand Euclidean distance, extending a fixed step length from the X_nearest to the X_rand direction, generating a new node X_new, detecting whether paths from the X_nearest to the X_new collide with obstacles or not and whether potential energy values of all discrete points on the paths meet potential energy threshold constraint or not, if the paths do not pass, discarding the extension, jumping to a step S2, and if the paths pass, entering the next step; S4, overall considering potential energy and path length to select an optimal father node, namely searching a potential father node set in a neighborhood with X_new as a circle center and setting a neighborhood searching radius, introducing a composite cost function comprising the path length and accumulated potential energy to calculate the total cost when each neighborhood node is used as the X_new father node, selecting the neighborhood node which enables the X_new to obtain the minimum cost as the father node, and adding the X_new into a random tree; S5, introducing rerouting optimization of potential energy weight, namely traversing other nodes in the neighborhood again after selecting a father node, judging whether the cost of reaching the neighborhood nodes through X_new is lower than the cost of reaching the neighborhood nodes without X_new, and changing the father node of the neighborhood nodes into X_new if the cost of reaching the neighborhood nodes through X_new is lower and no collision exists; s6, target judgment and path backtracking, namely judging whether the distance between the X_new and the target point X_ goal is smaller than a preset arrival threshold value or not: If not less than the preset reaching threshold, the iteration times are made Returning to the step S2 to continue execution; If the target point X_ goal is smaller than the preset arrival threshold value, the marker search is successful, the parent node index of each node is reversely traced back to the starting point X_start from the target point X_ goal, and a final low potential energy and collision-free safety path is extracted and output.
2. The global path planning method based on the potential energy guided fast search random tree algorithm according to claim 1, wherein the potential energy value of the position of the obstacle is recorded as 1, and the random point in the map does not comprise the obstacle.
3. The global path planning method based on potential energy guided fast search random tree algorithm according to claim 2, wherein any point in the map The potential energy value relative to each obstacle is calculated by: the method comprises the steps of obtaining, among others, Represents the value of the potential energy, For the rate of decay of the potential energy, Is taken as a point To an obstacle Is used for the distance of (a), In order to influence the range of the effect, Representing the distance from the point to the nearest point of the obstacle , ) Representing an obstacle The coordinates at the centroid of the sphere, Representing a square root taking operation.
4. The global path planning method based on the potential energy guided fast search random tree algorithm of claim 3, The range of the values is as follows 。
5. The global path planning method based on the potential energy guided fast search random tree algorithm according to claim 4, wherein the composite cost function comprising path length and accumulated potential energy is specifically: Wherein Representing the cost of planning the path and, Represents the potential energy value of the planned path, Representing the length of the planned path, And Is a coefficient and adds to 1.
6. An electronic device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of claim 5 when executing the computer program.
7. A computer readable storage medium storing computer instructions which, when executed by a processor, implement the steps of the method of claim 5.

Description

Global path planning method based on potential energy guiding rapid searching random tree algorithm Technical Field The invention belongs to the technical field of intelligent automobile path planning, and particularly relates to a global path planning method based on a potential energy guiding rapid search random tree algorithm. Background With the development of the age, intelligent automobiles have been provided with a prototype, and are expected to provide safe, comfortable and efficient driving environments for drivers in the future. Research on intelligent vehicles is mainly focused on improving safety and comfort of automobiles and providing an excellent human-vehicle interaction interface. In recent years, intelligent vehicles have become a hot spot of research in the world vehicle engineering field and a new power for the growth of the automobile industry, and many countries incorporate the intelligent vehicles into intelligent traffic systems which are each emphasized in development. In current global path planning algorithms, researchers tend to focus only on the behavior that an obstacle would interrupt the path connection, but ignoring the obstacle itself would also have an impact on the driver, who would want to be as far away from the obstacle area as possible, rather than "side-by-side". The above-mentioned behavior can ensure that the driver has enough time to correct errors and ensure the safety of the whole driving environment. Conventional fast search random tree algorithm (RRT) usually only aims at the single problem that an obstacle can block node connection, and often causes a planned track to run close to the obstacle, which is not consistent with driving practice. Disclosure of Invention In order to solve the single problem that the conventional rapid search random tree algorithm (RRT) usually only aims at blocking node connection when an obstacle is noticed, and the planned track is usually caused to run close to the obstacle, which does not conform to the technical problem of driving practical conditions, the application provides a global path planning method based on the potential energy guiding rapid search random tree algorithm. Firstly, defining the position of an obstacle and the position of a starting point and an ending point according to a two-dimensional space map planned by a vehicle, and aiming at any point in the mapCalculating potential energy values of the obstacle relative to each obstacle, and superposing the potential energy values of the obstacle relative to all the obstacles to obtain any pointAccording to the same method, potential energy values of all arbitrary points in a two-dimensional space map are obtained, the potential energy value of each point of the whole map is normalized, the normalized numerical value is between 0 and 1, a potential energy map matrix is obtained, the potential energy map matrix is introduced into a fast search random tree algorithm, a potential energy threshold value is set in an initial stage of path planning, randomly generated points are screened, the potential energy value is introduced into a cost function of the fast search random tree algorithm after screening, and wiring optimization is carried out again, so that vehicle path planning is carried out. Further, any point in the map does not include an obstacle, and the potential energy value of the position of the obstacle is recorded as 1. Further, any point in the mapThe potential energy value relative to each obstacle is calculated by: the method comprises the steps of obtaining, among others, Represents the value of the potential energy,For the rate of decay of the potential energy,Is taken as a pointTo an obstacleIs used for the distance of (a),In order to influence the range of the effect,Representing the distance from the point to the nearest point of the obstacle,) Representing an obstacleThe coordinates at the centroid of the sphere,Representing a square root taking operation. Further, the method comprises the steps of,The range of the values is as follows。 Further, the path planning method specifically comprises the following steps: s1, initializing a two-dimensional space map of vehicle planning, setting a starting point X_start and a target point X_ goal, loading barrier information, simultaneously importing an environment potential energy map matrix, and setting a maximum iteration number K, an expansion step length, a neighborhood search radius and a potential energy safety threshold value ; S2, constraint sampling based on potential energy guidance, namely randomly sampling in a map space to generate a candidate point X_rand, and reading a potential energy value corresponding to the point in a potential energy mapJudgingWhether or not it is less than a set potential energy threshold: If it isJudging that the point is in a high risk area, directly discarding the sampling point and re-sampling; If it is Reserving the point as an effective sampling point and ente