Research on Path Planning of Mobile Robot Based on DWA-IMP-A* Algorithm

Path planning is one of the most critical links in mobile robots. Its timeliness, security and accessibility are crucial to the development and wide application of mobile robots. However, in solving the problem of path planning, the most popular A* algorithm has some problems, such as heuristic function cannot be estimated accurately, node redundancy, path is not smooth, and obstacle avoidance cannot be achieved in real time. To solve these problems, A fusion algorithm of improved A* combined with reverse path and dynamic window method (DWA-IMP-A*) was proposed. The algorithm refines the heuristic function by incorporating the reverse path. The node optimization algorithm is used to further reduce the path length. The generated trajectories are smoothed by cubic spline interpolation. At the same time, it is integrated with the improved DWA algorithm to improve the efficiency and safety of robot path planning. The algorithm takes ROS mobile robot as the carrier and is tested under typical road conditions. Compared with A* algorithm, the planning time is reduced by 54.6% and the path length is reduced by 6.37%. Experimental results verify the effectiveness and robustness of the algorithm. The research results have certain reference significance for the path planning of various types of mobile robots and the research of driverless vehicles.