Decentralized control of robot manipulators: nonlinear and adaptive approaches

For robot arm tracking, the authors propose a nonlinear and adaptive approach based on decentralized system structure. Using the passive feature of robots and cubic feedback to treat the nonlinear couplings and quadratic interconnections, the decentralized adaptation is achieved by applying the linear-in-parameters property of the motion equation. The nonlinear feedback improves the performance of PD control from local to global stability and the adaptation reduces its tracking errors. The practical significance of the approaches lies in the fact that they can be implemented in most robots without hardware alteration.