Robot path planning based on improved A* algorithm

摘要

The A* algorithm has been widely employed in the automatic navigation of mobile robots. However, the traditional A* algorithm faces challenges in handling complex environments and large-scale maps due to high computational complexity, large memory requirements, and suboptimal performance under real-time constraints. To address these issues and enhance search efficiency, reduce memory usage, and improve path quality, an improved A* algorithm based on a 24-neighborhood search is proposed. Firstly, a safety strategy is introduced to improve path safety by setting appropriate buffer zones and disabling unsafe search. Secondly, the 24-neighborhood is reduced to eight search intervals according to the orientation of the destination, which reduces the number of turning points and improves search speed. Then, the evaluation function is improved by adding an exponentially weighted turning cost, which speeds up the search process and reduces the total turning angle. Finally, the modified A* algorithm is simulated multiple times using Python, and the results show improvements in planning time, number of turning points, path safety, and total turning angle.