Decentralized adaptive control for robot arm tracking

We propose a decentralized (or so called independent-joint) adaptive control scheme. The adaptive control is developed based on the passivity and linear-in-parameters properties of the robot motion equation. The proposed scheme, which consists of a PD term, a nonlinear term and an adaptation term, ensures the globally ultimately bounded tracking errors and parameter estimation errors. The practical significance of the scheme lies in the fact that it can be implemented in most robot manipulators without hardware alteration.