基于平滑蚁群算法的机器人路径规划

Mobile robot path planning based on smoothing ant colony algorithm

在栅格环境下利用蚁群算法规划出来的移动机器人路径存在转弯次数多,累计转折角大等问题。针对这些问题,提出了平滑蚁群算法。平滑蚁群算法是在蚁群算法规划出来的最优路径的基础上,将当前节点与其它不在同一条直线上的节点依次连线,如果新的连接线不穿越障碍物,则将当前连接线作为新路径代替原来路径,并删除中间节点,否则对当前路径不作任何修改。仿真结果表明,平滑蚁群算法能够在不同栅格规模和障碍物覆盖率不同的情况下,有效降低路径长度,减少转折次数和累计转折角度,并且路经规划结果优于免疫遗传算法和势场蚁群算法的路径规划结果。

Path planning for mobile robot by ant colony algorithm is flaw with frequently turning points,large cumulative turning angle under grid environment.Smoothing ant colony algorithm is proposed to solve these problems.Smoothing ant colony algorithm is based on the optimal path planned by ant colony algorithm,connecting the current node and the nodes after the turning point,deleting original node and replacing the initial path by the new line as no obstacles existing on the line.Otherwise,the current path is not modified.The simulation results show that smoothing ant colony algorithm exceeds immune genetic algorithm and ant colony optimization and artificial potential.Length,total turning points and cumulative turning angle of path are almost reduced when smoothing ant colony algorithm is adopted.Path planning problem under different complex environment with different obstacles distribution can be achieved by smoothing ant colony algorithm.