Path Planning Based on Fusion of Improved A* and DWA Algorithm

Obstacles are often unevenly distributed in the sea area. For example, obstacles are very rare in the open sea area, while a large number of obstacles gather in the coastal area. This leads to A large number of redundant expansion points in the exploration process of traditional A* algorithm. To solve this problem, this paper proposes an improved algorithm and integrates dynamic window approach (DWA) to realize dynamic obstacle avoidance. Firstly, by calculating the proportion of local obstacles in real time, an adaptive neighborhood search strategy is selected: jump points are used to expand the neighborhood in sparse areas, and 8 neighborhood is used to expand the neighborhood in dense areas, effectively improving the search efficiency. Secondly, in view of the hidden danger that the optimized path may touch the obstacle vertices, a collision safety detection mechanism is introduced in the process of eliminating redundant points to ensure that the path maintains a safe distance from the obstacle. Finally, the improved A algorithm is integrated with DWA algorithm, and the local dynamic obstacle avoidance ability of DWA is used to correct the global path in real time. Comparative experiments were carried out on the MATLAB simulation platform, which showed that the improved algorithm was optimized in terms of track length, track smoothness and elapsed time.