Global path planning for mobile robots based on improved A-Star algorithm

Abstract

Aiming at the problems of lengthy planning time, easy to fall into local optimization, and insufficient ability to adapt to complex environments in the global path planning process of A-Star algorithm, we propose an improved A-Star algorithm. First, we introduce global obstacle information and local obstacle information into the heuristic function to adapt to different environmental requirements, and propose a U-shaped "trap" region filling strategy to reduce the number of searched nodes and improve the search efficiency. Secondly, we assign a priority to the search neighborhood and use directional gain to enhance the orientation of the search process. Then, we introduce a key node selection strategy to eliminate redundant nodes on the path and improve the smoothness of the path. Simulation results prove that the algorithm in this paper has significant improvement in terms of the number of traversed nodes, path nodes, path inflection nodes, turning angle, pathfinding time and path length. Finally, we experiment with the improved algorithm in a real environment, proving the effectiveness of the improved algorithm and demonstrating its ability to significantly enhance the performance of mobile robot path planning.