基于稀疏节点快速扩展随机树的移动机械臂运动规划

Motion Planning of Mobile Manipulators Based on RRT with Sparse Nodes

针对传统移动机械臂运动规划未考虑停靠误差、规划效率低等问题,提出一种基于稀疏节点快速扩展随机树的移动机械臂运动规划方法。研究了停靠误差对移动机械臂运动规划的影响,制定了基于坐标转换的误差补偿措施。针对快速扩展随机树算法在局部空间中过度搜索的问题,改进回归过滤机制,避免了节点重复扩展。考虑边界节点的有效性,提出了随机点边界扩展机制以缩短碰撞检测时间。实验结果表明所提方法提高了运动规划效率。

Traditional motion planning of mobile manipulators had problems such as without considering docking errors and low efficiency.A novel motion planning method of mobile manipulators was proposed based on RRT with sparse nodes.Influences of positioning errors for the mobile platforms on motion planning were analyzed,and an error compensation method was developed based on coordinate transformation.Aiming at addressing excessive searching in local spaces by RRT algorithm,a regression filtering mechanism was improved to avoid repeated expansion of nodes.Then,considering validity of boundary nodes,the random point boundary extension mechanism was proposed to greatly decrease collision detection time.Experimental results demonstrate that the proposed algorithm improves efficiency of motion planning.