Path Planning for Unmanned Driving Based on Improved A* Algorithm

The A* algorithm is currently the most commonly used basic algorithm and is used in various fields. In order to make its performance better and more suitable for the vehicle kinematics trajectory, it is specially optimized. In this paper, the octile function is used as the extended heuristic function of 8 adjacent nodes, and the shake-free function is added to reduce the influence of the ideal path on the path planning results, and the search performance is greatly improved by adding weight to it. The artificial potential field method is used to assign the cost to obstacles and surrounding grids, effectively avoiding the edge optimal phenomenon. And finally uses the Bezier curve to make the planning a result more suitable for the vehicle's driving path.