基于改进APF-Bi-RRT*算法的移动机器人路径规划研究OA

To address the issues of sampling point redundancy,low-quality path planning,and excessive path cur-vature in the Bidirectional Rapidly-exploring Random Tree(Bi-RRT*)algorithm for mobile robot path planning,an improved APF-Bi-RRT* algorithm is proposed.First,a dynamic target bias function is introduced into the Bi-RRT*to reduce the number of sampling points and improve the path planning efficiency.Second,a dynamic repulsion co-efficient is incorporated into the Artificial Potential Field(APF)to enable real-time adjustment of the repulsive force,thereby enhancing the quality of the path planning.Finally,the pruning algorithm is improved by integrating a safe distance detection strategy combined with cubic B-spline curves to alleviate the path curvature.The proposed algorithm was validated through simulation experiments in diverse environments.The results show that it improves planning efficiency by 16.46％,22.52％,and 62.68％compared to the APF-Bi-RRT*,Bi-RRT*,and RRT* algo-rithms,respectively.The improved algorithm significantly reduces computation time and the number of sampling points,while achieving shorter path lengths and higher smoothness.