Unmanned Vehicle Global Planning Based on Improved A* Algorithm
Kai Xin, Huacai Lu · 2025
Aiming at the problems of excessive search nodes, low efficiency, and node redundancy in the traditional $A^{*}$ algorithm for global path planning of unmanned vehicles on unstructured roads, a method to optimize the heuristic function is proposed. The optimization approach in this paper involves adding a weight to the estimation function, making the estimated value of the estimation function infinitely close to the real cost value. This allows grid nodes in less probable directions to be removed from the search scope when filtering nodes during traversal. Through simulation experiments using PyCharm software, the effects of the traditional $A^{*}$ algorithm and the weighted distance $A^{*}$ algorithm were compared in scenarios with the same obstacle settings. The experimental results show that compared with the traditional A* algorithm, the improved A* algorithm has significant advantages in terms of expanded nodes, search time, and the number of turns, effectively solving the global path planning problem for unmanned vehicles on unstructured roads.