A novel task-space based motion planning method for redundant manipulators with trap detection
摘要
Most heuristic approaches for redundant manipulators provide reliable theoretical assurance such as completeness and bounds on sub-optimality. However, dealing with the high-dimensional planning problem in the task space by heuristic search is still challenging. On the one hand, the path planning of some redundant manipulators occasionally falls into traps, which could lead to planning failure. Besides, occasional discontinuous jumps in the joint space paths may cause a collision between a manipulator and surrounding obstacles. Therefore, a heuristic search-based motion planning method is proposed to solve this problem, in which a trap detection algorithm is employed to check the effectiveness of joint space paths, and a predecessor detection algorithm is designed to prevent joint paths from jumping. Experimental results show that the proposed motion planning method could generate consistent, low-cost motion trajectories and deal with the problem of cluttered spatial path planning.