Constrained Path Planning for Reconfiguration of Cable-Driven Parallel Robots

摘要

Obstacle avoidance is a challenging problem in cable-driven parallel robots (CDPRs) due to the cross-distribution of cables in the workspace. To address this challenge, a novel constrained path planning method is proposed in this article to realize the reconfiguration of CDPRs, where the cable anchor positions and cable lengths are adjusted simultaneously in the reconfiguration to achieve obstacle avoidance. To realize such reconfiguration, the loop-closure constraints are first formulated and the constrained configuration space is established in terms of the implicitly represented manifold. Then, combined with the designed wrench feasible detection and collision detection that are responsible for the positive cable tension guarantee and obstacle avoidance, the bidirectional rapidly exploring random trees are utilized to generate branches in the tangent spaces of the manifold to finally discover a feasible path in the whole constrained configuration space. Finally, simulations and experiments are implemented on the 3-DOF reconfigurable CDPR and the results indicate that the proposed method can coordinate the changes of the cable anchor positions and cable lengths by using the loop-closure constraints, and a collision-free path during the reconfiguration can be successfully found in the presence of regular or irregular obstacles.