Mobile robot path planning based on improved A-star_DWA fusion algorithm
Yong Li, Sen Qi, Chaoxing Zhang · 2022
In mobile robot navigation, a single A-star algorithm or Dynamic Window Approach (DWA) algorithm cannot combine global path optimization and real-time obstacle avoidance. In this paper, we propose a robot navigation algorithm that improves the fusion of A-star algorithm and DWA algorithm. To address the problems of too large search range of the A-star algorithm and the existence of redundant path inflection points in the planned paths, a coefficient representing the environmental obstacle rate is first introduced to the cost function of A-star to improve the search efficiency; Secondly, a key point selection strategy is proposed to remove the redundant points in the path and retain only the necessary key points to improve the smoothness of the path. Then, the evaluation function of the DWA algorithm is constructed by combining the global path key point information, and the key points are used as intermediate target points of the DWA algorithm to plan local paths, improve the smoothness of paths, and realize the real-time obstacle avoidance function on the basis of achieving the optimal global path. Finally, the effectiveness and feasibility of the proposed navigation algorithm are verified by simulation experiments.