Adaptive linearization for cooperative position control of robotic manipulators

Abstract

Multilink manipulators have nonlinear and strongly coupled dynamics, which can be linearized and decoupled by using nonlinear feedback control schemes. Since the nonlinear feedback law depends on the system's parameters, if the system parameters change, the linearized system will go beck to a nonlinear system. This paper shows how the adaptive control technique can be used to overcome system parameter variation. The steepest descent method is used to adjust the parameters in the nonlinear feedback loop. Once the system is linearized adaptively, the controller for cooperative position control can be designed using linear control techniques.<>