Robust control of robot manipulators based on disturbance observer in task space

Abstract

A robust and high response hybrid position/force control strategy based on controlling the second derivatives of control variables in task space such as position and force is proposed. A position-based disturbance observer and a force-based disturbance observer in task space are used to control the second derivatives of position and force, respectively. Experiments are performed by a direct drive robot to show the effectiveness of the proposed strategy. In this paper, two types of inverse kinematics (inverse Jacobian matrix and transposed Jacobian matrix) are tested and compared each other. The analysis and the experiments show that there is a trade-off between the performance and the computation effort of two inverse kinematics.<>