关键词: 车辆工程； A*算法；路径规划；自动驾驶

Abstract: When using the traditional A* algorithm for path planning, there are problems such as too many turning points, poor search efficiency and too large turning angle. Therefore, a path planning algorithm for autonomously driving vehicles based on the adaptive A* algorithm was proposed. First of all, in view of the long operation time of the traditional A* algorithm, the exponential growth method was used to weight the heuristic function, so as to improve the search efficiency. Then, the heading angle influence factor was introduced into the heuristic function, so that autonomously driving vehicles could meet the steering characteristics when working. Finally, through the cubic B-spline interpolation function, the inflection point in the path was smoothed, so that the local path with sharp corners at the inflection point is smoother. The MATLAB simulation results show that the performance indexes of the path generated by the improved A* algorithm are better than those of the traditional A* algorithm in terms of length, total heading angle and curve smoothness, so that the autonomously driving vehicle can obtain a high-quality path that satisfies its own constraints. Moreover, under different map environments, the proposed adaptive A* algorithm can adapt to different environments and can effectively maintain its feasibility and superiority.

Key words: vehicle engineering; A* algorithm; path planning; autopilot