Passivity based hierarchical multi-task tracking control for redundant manipulators with uncertainties

Abstract

Most research so far on hierarchical multi-task control with strict priority requires exact dynamic and kinematic models. However, when a robot picks up tools of unknown lengths, orientations and mass, its kinematics and dynamics are varying and unavailable. There are few controllers that can achieve asymptotic tracking of multi-task in the presence of kinematic and dynamic uncertainties. Thus, in this paper we present a passivity-based adaptive Jacobian hierarchical multi-task controller. New reference velocity and kinematic regressor matrix are defined for the multi-task case with strict priority to guarantee the asymptotic convergence of both tracking error and parameter estimation error. A novel natural adaption law is specially designed for this method to guarantee the correctness of estimated dynamic parameters and increase the rate of convergence. Experiment results are presented to illustrate the performance of the proposed controller.