基于改进RRT-Connect算法的机械臂路径规划

针对RRT^*算法速度较慢问题,提出一种快速收敛至最优路径的最优双向快速扩展随机树(Optimal Bidirectional Rapidly-exploring Random Trees,Obi-RRT)算法。Obi-RRT使用改进的RRT-Connect算法快速得到较低成本路径,通过路径修剪得到关键点,围绕关键点提出三种采样空间并进行采样,通过不断更新关键点从而得到最优或接近最优的路径。平面和机械臂关节空间下的仿真实验表明,Obi-RRT算法运行时间仅为RRT^*算法的十分之一,并且路径成本更低。

PATH PLANNING OF ROBOT ARM BASED ON IMPROVED RRT ALGORITHM

Aiming at the problem of the low speed of RRT^*algorithm,this paper proposes an Obi-RRT algorithm that quickly converges to the optimal path.Obi-RRT used the improved RRT-Connect algorithm to quickly obtain a lower cost path,and obtained key points through path pruning.Three sampling spaces were proposed around the key points for sampling,and the key points were continuously updated to obtain an optimal or near optimal path.The simulation experiments in the plane and the joint space of the robot arm show that the running time of the Obi-RRT algorithm is only one tenth of the RRT^*algorithm,and the path cost is lower.