Jump point search path planning method in dynamic scene

Abstract

The invention aims to solve the problems of low algorithm efficiency, local optimal path and the like in a search-based path-finding algorithm in a rapidly changing dynamic complex scene. On the basis of a jump point search (JPS) algorithm, a dynamic jump point search (DJPS) path planning method in a dynamic scene is provided, and on the basis of the JPS algorithm, the dynamic jump point search (DJPS) path planning method in the dynamic scene is provided. The DJPS algorithm is used for coping with obstacle changes in different scenes, a set of complete'jump point-path updating scheme 'is designed, meanwhile, a new'near path' possibly occurring during obstacle updating is explored, and the obtained path is optimized to ensure the optimality of the path. In order to verify the effectiveness of the DJPS algorithm, path updating experiments under various complex maps are designed. According to the method, obstacle detection is only performed on the original path and the nodes near the path, and the search length during path finding each time is limited by utilizing jump points stored in OpenList and CloseList in the original algorithm and'abandoned paths' which are obtained by mutual connection and do not finish path finding, so that the time complexity and the space complexity of the algorithm are reduced, and the path finding efficiency is improved. And a jump point searching scheme in a dynamic scene is realized.

Description

technical field [0001] The present invention relates to the field of artificial intelligence and unmanned systems, in particular to the technical field of automatic driving path planning, and in particular to a jump-point search path planning method in a dynamic scene. Background technique [0002] In the automatic driving of unmanned vehicles in the field of unmanned systems, path planning is one of the key technologies. A lot of research work has been carried out so far for this technology. Path planning for unmanned vehicles can be divided into static path planning and dynamic path planning. Among them, the A* algorithm and Dijkstra algorithm are mainly used in known static scenes, while the artificial potential field method and D* algorithm are mostly used in dynamic scenes with unknown environments. Among them, Jump Point Search (JumpPointSearch, JPS) is an optimization of the A* search algorithm. Through the pruning rules, the symmetry of the grid is broken, the acces...

AI Technical Summary

Problems solved by technology

[0007] (4) The "re-search strategy" proposed in this article will re-explore the JPS to the end point based on the current robot's motion position, which means that the path obtained in the last path-finding exploration is completely abandoned, and a large number of useless nodes are re-examined. scanning, which greatly reduces the computational efficiency of the algorithm in the face of complex scenes;

If the back-end path changes greatly in the next path that the robot is going to take, which leads to the failure to make a timely decision when the robot's movement route should be re-corrected, this will cause the robot to keep walking "wrong" Road", keep falling into "local optimum"

And this detection method does not take advantage of the advantages of quickly detecting a large number of nodes when the grid method is used for detection, and the detection nodes are limited to the vicinity of the robot position

The operating efficiency is also inferior to the artificial potential field method, and cannot handle dynamic map changes in fast and complex scenes

Images