改进RRT算法的机械臂路径规划

Improved RRT Algorithm on Manipulator Path Planning

针对传统快速扩展随机树(RRT)算法在机械臂避障路径规划时存在搜索效率低、随机性大的问题,提出一种改进RRT算法。该算法结合双向RRT算法和目标偏置RRT算法的优点,并在随机树扩展过程中设置虚拟目标点和改变步长。由于RRT算法规划的路径存在冗余节点多的问题,利用双向删除取最短路径的方法对避障路径中的冗余节点进行删除,最后利用三次B样条曲线对避障路径进行平滑处理。仿真结果表明,改进RRT算法在扩展时间、路径长度和扩展节点数均优于双向RRT算法和目标偏置RRT算法,对路径中冗余节点删除和平滑处理的有效性。

Aiming at the problem of low search efficiency and large randomness in the obstacle avoidance path planning of the traditional rapid expansion random tree(RRT)algorithm,an improved RRT algorithm is proposed.The algorithm combines the advantages of the bidirectional RRT algorithm and the target bias RRT algorithm,and sets the virtual target point and changes the step length during the random tree expansion process.Because the path planned by the RRT algorithm has many redundant nodes,the redundant nodes in the obstacle avoidance path are deleted by the method of two-way deletion and the shortest path,and finally the cubic B-spline curve is used to smooth the obstacle avoidance path.The simulation results show that the improved RRT algorithm is better than the two-way RRT algorithm and the target bias RRT algorithm in the expansion time,path length and number of expansion nodes,and is effective in deleting and smoothing redundant nodes in the path.