Application Of A * Algorithm Combined With Improved Artificial Potential Field In Robot Global Path Planning

Directly influences the performance of mobile robot path planning algorithm of work efficiency,thus optimizing the performance of path planning algorithm is especially important,this research mainly through the A * algorithm and artificial potential field method,the A * algorithm and artificial potential field method of existing limitations and pathfinding process analysis and optimization,algorithm and the improved fusion in this paper,A new global path planning algorithm,In this study,the proposed algorithm is analyzed theoretically and simulated by MATLAB.Finally,the proposed algorithm is transplanted to a robot for verification in an indoor environment.The proposed algorithm is mainly divided into three parts: map initialization,path planning and path smoothing.(1)In the actual path planning process,the disorderly distribution of obstacles may induce mobile robots to enter the interior of some concave obstacles to conduct path search,which will increase the search time and reduce the efficiency of path planning.Therefore,this study proposes a method to deal with some obstacles in the grid map,which can selectively deal with all obstacles on the whole map and the obstacles encountered in the direction of travel.After each obstacle in a local area is processed,the next obstacle to be dealt with will be re-planned.The algorithm uses a rectangular area to close all irregular obstacles in the local area,and does nothing to the obstacles outside the forward direction to avoid meaningless operations.Simulation analysis of the map preprocessing algorithm in MATLAB shows that the number of search nodes of A* algorithm combined with the map preprocessing algorithm is reduced by 12.12%～46.77% and the search time is accelerated by7.62%～46.04%,which proves the effectiveness and feasibility of the proposed raster map preprocessing algorithm.(2)In order to optimize the A * algorithm and artificial potential field method in path search easy generation of low efficiency and produce the problem such as local minimum value region,this study proposes A fusion of artificial potential field method three neighborhood search A * algorithm,the algorithm is mainly divided into three neighborhood search A * algorithm of precise and part of the artificial potential field method of obstacle avoidance dominant pathfinding two parts quickly,The A* algorithm of eight neighborhood search is optimized to three neighborhood search,which is combined with the artificial potential field method and the raster map initialization method.This algorithm can without obstacles blank areas are part of the artificial potential field method for fast search path,area near the obstacles is the switch for three neighborhood search precision A * algorithm of obstacle avoidance,through three neighborhood search A * algorithm and part of the artificial potential field method both pathfinding algorithm constantly switching and cooperate,to quickly build path curve can be achieved without collision.Simulation analysis of the proposed algorithm in MATLAB shows that,compared with A* algorithm,the path search time of A* algorithm based on artificial potential field method is reduced by 72.14%～81.39%,the path length is shortened by 3.86%～7.03%,and the number of path nodes is reduced by 86.96%～90.52%.It can smooth the path to some extent and reduce the turning point of the path.(3)In view of the smoothing of the path curve in the grid map,such as the existence of sharp points and insufficient security,a path smoothing optimization algorithm based on the local artificial potential field method is proposed,which selectively assigns a local artificial potential field method to a single obstacle by analyzing the distribution location information of the path turn and the obstacle,and rewires the original path curve to achieve the effect of smoothing the path,compared with the traditional path smoothing method,the continuity,feasibility in actual motion,There are better results in terms of security and path length.Compared with the Bézier curve smoothing method,this institute proposes that the algorithm can effectively optimize the initial path length,and the optimization ratio varies from 0.59% to 5.14%,and then the algorithm in this chapter can optimize the number of curves of the initial path from 22.22% to 66.66%,thereby improving the quality of the path curve.(4)The experimental platform was built,the algorithm of this study was verified by the robot in the indoor environment,and the experiments were carried out in the comprehensive scenario,the narrow and long aisle scene and the barrier-free scene,and the results of the experiment and the causes were theoretically analyzed,which proved that the algorithm of the study could plan a better path,optimize the path length,and be more suitable for the actual movement of the mobile robot than the A* algorithm.