Mobile Robot Path Planning Method Based on Improved A* Algorithm and Dynamic Window Approach in Complex Environments

For tackling the challenges in robot path planning within complex environments, such as low computational efficiency, excessive redundant nodes, poor obstacle adaptability, high collision risk, and lack of global guidance, an enhanced path planning method integrating A* and dynamic window approach (DWA) is developed. The obstacle distance information is introduced as a risk cost into the heuristic function, enabling the planned path to stay farther away from obstacles, avoid entering obstacle-dense areas, and improve path safety and feasibility. The dynamic weight factor is introduced to reduce invalid node expansions and enhance search efficiency. A bidirectional Floyd path post-processing algorithm is implemented to eliminate redundant nodes and inflection points. The global path key points are extracted and utilized as guiding intermediate points for the local planning approach, and the DWA evaluation function is improved to enhance path smoothness and avoid local optima. Simulation results verify that the algorithm exhibits better global optimality and stronger safety.