Gradient Projection of Weighted Jacobian Matrix Method for Inverse Kinematics of a Space Robot With a Controlled-Floating Base

Abstract

This paper studies the inverse kinematics (IKs) of a space robot with a controlled-floating base. Different from the traditional space robot which has a free-floating base, the momentum conservation is no longer satisfied so that the degrees-of-freedom (DOFs) and redundancy of the robot obviously increase, and motion limits exist for both base and manipulator. To deal with such a problem, a gradient projection of weighted Jacobian matrix (GPWJM) method is proposed. The Jacobian matrix is derived considering the additional DOFs of the base, and the trajectory tracking by the end-effector is chosen as the main task. A clamping weighted least norm scheme is introduced into the derived Jacobian matrix to avoid the motion limits, and the singular-robustness is enhanced by the damping least-squares. The convergence and accuracy analysis indicates the calculation of damping factor; while the verification of motion limits avoidance indicates the inequality constraint of clamping velocity. Finally, the effectiveness of the proposed GPWJM method is investigated by the numerical simulation in which a planar 3DOF manipulator on a 3DOF base is taken as a demo.