Path Tracking of a Cable-Driven Snake Robot With a Two-Level Motion Planning Method

摘要

Due to high degree of freedoms (DOFs), motion planning of the cable-driven snake robot is challenging and usually tackled by the iterative Jacobian-based method. This iterative method adopts discontinuous velocity and tolerance to determine intermediate poses, which deteriorate the tracking performance of robots. In this paper, a two-level motion planning method is proposed to schedule the motion of cable-driven snake robot owning a serial-parallel hybrid structure. First, kinematics model of the robot among actuators, joints, and end effectors are introduced. Second, in the upper level, a jerk-bounded algorithm is employed to plan the position, velocity, and acceleration profiles of the end effector for a given path, whereas in the lower level, the joint angles and cable length changes are calculated by the inverse kinematics of the robot according to the position and velocity profiles of the end effector from the upper level. Therefore, the iterative process is eliminated by the presented method, which enhances the tracking performance and motion smoothness of the robot. Finally, simulations and experiments on a 24 DOFs snake robot prototype demonstrate the effectiveness of the presented method. According to the experimental results, the two-level method improves the repeatability and accuracy of the prototype by about 60% and 6%, respectively, compared with the iterative method.