Globally fully adaptive decentralized control of robot manipulators

Abstract

We develop a globally fully adaptive decentralized controller of a robot manipulator for trajectory tracking. With some nonlinear feedback terms, the proposed decentralized control law is guaranteed to drive the tracking error asymptotically to zero. Moreover, those nonlinear feedback terms not only improve the closed-loop property from semi-global stability to global stability but also make the position and velocity tracking errors approach to zero, which is a dual-goal hardly attained in the literature. The proposed adaptive law allows all signals in the closed-loop systems to be stabilized while making the tracking error converge to zero, even without any prior knowledge of the robot manipulator or the payload and the bound on the design trajectory so that the law can be claimed to be fully adaptive. Finally, a numerical study is provided to verify the effectiveness of the proposed scheme.