Abstract: To address the problem that a single algorithm in path planning and navigation for mobile robots cannot satisfy both path optimization and real-time obstacle avoidance, an improved navigation algorithm that combines an improved A⁎ algorithm and an improved dynamic window approach (DWA) is proposed. The quantified obstacle information was used as the adaptive adjustment weight of the heuristic function in the A⁎ algorithm to improve the search efficiency of the algorithm, and the vector method was used to remove the common line nodes and the key point extraction method to remove the redundant turning points so as to improve the smoothness of the path. For DWA algorithm with unreachable targets and planning paths that did not fit with global paths, we proposed to dynamically adjust the azimuth angle and introduce the evaluation function of distance to the end position, and the path of improved algorithm was closer to the global path. The two algorithms were combined with key point information. Through comparison by simulation and experiment, it is shown that the improved A⁎ and DWA fusion algorithms perform well in both unknown static and dynamic environments.