改进A*算法的智能车路径规划研究

Research on Intelligent Vehicle Path Planning Based on Improved A*Algorithm

由于A*算法所规划的路径存在着转折次数多,路径不平滑,路径贴合障碍物和初始时刻转折角度过大等不符合车辆运动学的问题;为了解决上述问题,获得适用于智能车的优化路径,首先通过对车辆运动学进行建模得到车辆的约束,同时在估价函数中加入车身轮廓代价和障碍物距离代价,并将车辆约束加入到A*算法的启发函数和路径优化中,最后使用贝塞尔曲线拟合转折点,使A*算法所生成的路径更加符合车辆的运动学;实验结果表明,改进后的算法所规划的路径更加平滑、合理,且符合车辆的运动特性。

There are several problems existing in the path planned by A*algorithm in real traffic situations.These problems included too many turning points,unsmooth path,path fitting roadside and large turning angle at the beginning,which is not satisfied with the kinematic model of the vehicle.In order to solve the above problems and obtain an optimized path suitable for smart cars,the vehicle constraints are obtained by modeling the vehicle kinematics firstly.Then body contour cost of the car and the obstacle distance cost are added to the evaluation function.At last,the vehicle constraints are added to A*algorithm's heuristic function and path optimization,and Bezier curves are used to fit the turning points so that the path generated by the A*algorithm is more suitable for the kinematics of the vehicle.The results of improved A*algorithm demonstrated that the path planned by the improved algorithm is smoother,more reasonable and suitable for motion characteristics of the vehicle.