To address the issues of low search efficiency, excessive node expansion, and the presence of redundant nodes in the traditional A* algorithm, this article proposes an improved A* algorithm for mobile robot path planning. Firstly, a multi-neighborhood hybrid search method is introduced, optimizing the traditional eight-neighborhood and twenty-four-neighborhood into a new sixteen-neighborhood. The choice between eight-neighborhood search and sixteen-neighborhood search is determined based on the presence of obstacles in the eight-neighborhood around the current node, effectively enhancing the search efficiency of the algorithm and reducing the number of nodes expanded during the search process. Subsequently, unnecessary nodes are eliminated based on the positional relationship between the current node and the target node, according to neighborhood direction search rules, further decreasing the number of expanded nodes. Additionally, improvements to the bidirectional search mechanism along with the incorporation of dynamic weight coefficients further enhance the search efficiency of the algorithm. Furthermore, a strategy for extracting key nodes is employed to effectively remove useless turn points, thus resolving the issue of redundant nodes. Finally, simulation experiments demonstrate that the proposed improved A* algorithm outperforms the traditional A* algorithm in terms of search speed, number of expanded nodes, and path length, validating the effectiveness of the proposed method.