Research on Mobile Robot Path Planning Based on Improved A* and DWA Algorithms

This paper explores the navigation of robots, which involves two key aspects: global path planning and local path planning. The current approach utilizes the A* algorithm for global planning and the DWA algorithm for local planning. However, the traditional A* algorithm often fails to consider obstacles, leading to impractical paths that robots struggle to navigate. As a result, success rates are narrow. To address these issues, we propose an enhanced A* algorithm that optimizes the search approach, heuristic function as well as path smoothing. This modification ensures that the generated paths align better with the robot's motion while still prioritizing the shortest distance. Additionally, we refine the evaluation function of the DWA algorithm to account for the robot's angular velocity at different linear velocities. This adjustment enables smoother steering through curves while maintaining a consistent linear velocity at the same time. The improved algorithm greatly improves the speed of the robot when cornering and has a more reasonable completion path, while the success rate of cornering is greatly improved. In our physical map, the robot with the improved algorithm consumes 33.95% less time than the robot with the traditional algorithm, and the robot's path is much better.