A*算法在AGV路径规划上的改进与验证

针对单AGV路径规划时,A^*算法的启发函数采用曼哈顿距离时遇到障碍物会出现局部绕行这一问题,将带有障碍物的栅格地图作为环境模型,研究出两种改进A^*算法的路径规划方法。第一种方法是在遇到障碍物时将启发函数中曼哈顿距离换成欧氏距离,利用欧氏距离规划路径代价最小的特性避免绕行;第二种方法通过比较AGV遇到障碍物的位置与障碍物左右两端距离大小,通过规定行驶方向避免绕行。仿真结果表明,两种方法均可以在单AGV遇到障碍物时避免绕行,有效地减少了行驶时间,也使得路径更加平滑,提高了AGV的运行效率。

IMPROVEMENT AND VERIFICATION OF A*ALGORITHM IN AGV PATH PLANNING

Aiming at the problem of local detour of obstacles when using Manhattan distance as heuristic function of A^*algorithm in single AGV path planning,taking grid map with obstacles as environment model,two path planning methods of improved A^*algorithm are studied.The first method is to change the Manhattan distance of heuristic function to Euclidean distance when encountering obstacles,and the least cost of Euclidean distance planning is used to avoid detour.The second method is to compare the distance between the location of the obstacles encountered by AGV and the left and right ends of the obstacles,and to avoid detour by specifying the driving direction.The simulation results show that the two methods can avoid detour when single AGV encounters obstacles,which reduces effectively the driving time,makes the path smoother,and improves the operation efficiency of AGV.