Pose-Shape Unified Planning of Cable-driven Serial Manipulators based on Multiple Arm-Angle Construction

Abstract

Cable-driven serial manipulator (CDSM) has slender body and high dexterity. Hence it has large potentials in confined space application. However, in terms of complicated motion, achieving overall shape planning and control for the manipulator as a whole has been rarely accomplished, greatly limiting the utilization of redundant degrees of freedom. Therefore, this paper proposes a pose-shape unified planning method based on multiple arm-angle construction and optimization. Firstly, we derive the multi-space mapping relationships, in particularly introducing the concept of arm shape for joint space and deriving the corresponding analytical expressions. Subsequently, we derive the Jacobian matrix of Multiple arm-angel, which theoretically guarantees the efficiency of planning and the ability to achieve acceptable goals in a globally feasible and stable manner. Simulations demonstrate that our framework enables rapid and efficient arm-angles planning while ensuring high pose accuracy. Finally, experiments are conducted using the CDSM to verify the arm's motion capabilities, including pose adjustment and maintaining arm shape during movement. Compared to other planning and closed-loop control methods, the results confirm the consistency and reliability of the arm shape and end-effector pose.