An Improved Path Planning Algorithm Based on A* Algorithm

Path planning is a crucial aspect of vehicle navigation, and this paper presents an enhancement to the classic A* algorithm to address key challenges in this domain. The proposed method aims to improve both the efficiency and safety of path planning. In practical applications, path planning encounters various issues, such as an excessive number of unnecessary nodes during the search process, resulting in suboptimal planning efficiency. Additionally, obstacles may be present along the route between the starting point and the target node, requiring obstacle path search. Moreover, traditional cost functions often fail to fully account for vehicle safety, thereby increasing the risk of collisions. To overcome these challenges, the proposed method incorporates two key enhancements. Firstly, it employs a node marking technique on the grid map to identify key nodes and reduce the search process for unnecessary nodes, thus enhancing planning efficiency. Secondly, an incremental expansion of search nodes is utilized, employing an improved A* algorithm with a modified cost function. This enables the algorithm to plan collision-free paths from the starting point to the key nodes while considering the distance cost associated with potential collisions, thereby enhancing vehicle safety. Experimental results demonstrate that the proposed method significantly enhances path planning efficiency by reducing the search efforts for unnecessary nodes, resulting in accelerated path planning. Furthermore, the improved cost function enables the generation of safe and feasible paths, thereby enhancing vehicle safety.