基于改进双向RRT^(*)算法的机械臂路径规划

Path planning of manipulator based on improved two-way RRT^(*) algorithm

针对传统双向改进快速扩展随机树(RRT^(*))算法采样速度较慢以及产生路径较粗糙的问题,对6轴机械臂的运动规划进行分析,提出了一种基于目标采样和局部路径优化改进的双向RRT^(*)算法。首先,应用目标采样思想,在对扩展点进行筛选时,考虑随机扩展点与目标点的距离,保留距离目标点最近的随机搜索点;然后,对该搜索点进行扩展,减少不必要区域的搜索,使采样更加高效,搜索路径更优;最后,在此基础上,结合局部路径优化算法,对RRT^(*)算法中重布线随机树过程引入优化函数概念,计算两个节点之间最小距离所需节点,并连接该节点以降低路径曲折程度。将所提改进算法与原始双向RRT^(*)算法在不同测试环境下用机器人操作系统(ROS)进行仿真实验,实验结果表明,所提算法能有效优化算法规划路径,并提高规划效率。

Aiming at the slow sampling speed and rough path of the traditional two-way improved Rapidly-exploring Random Tree(RRT^(*))algorithm,the motion planning of 6-axis manipulator was analyzed,and an improved two-way RRT^(*)algorithm based on target sampling and local path optimization was proposed.Firstly,using the idea of target sampling,when screening the extension points,the distances between the random extension points and the target points were considered,and the nearest random search point from the target point was retained.Then,the search point was extended to reduce the search of unnecessary areas,so that the sampling was more efficient and the search path was better.Finally,on this basis,combined with the local path optimization algorithm,the concept of optimization function was introduced to the rerouting random tree process in the RRT^(*)algorithm,the node required for the minimum distance between two nodes was calculated,and the node was connected to reduce the path tortuosity.The proposed improved algorithm and the original two-way RRT^(*)algorithm were simulated with Robot Operation System(ROS)under different test environments.Experimental results show that,the proposed algorithm can effectively optimize the algorithm planned path and improve the planning efficiency.