Indirect adaptive control for nonlinear robotic systems

This paper presents an indirect adaptive control scheme for trajectory tracking of robotic manipulators. For each joint, the system parameters are estimated online using a modified recursive least squares algorithm with data normalization. A variable dead zone is incorporated into this estimation algorithm to limit the effects of interconnections and nonlinear dynamics in the system such that the number of the parameters to be estimated is reduced. The adaptive control law consisting of a pole placement PID controller and a nonlinear feedback is calculated based on current system parameter estimates. The nonlinear dynamics of the system is compensated by the adaptive nonlinear feedback. The proposed adaptive control system is shown to be globally stable and is implemented on a transputer network for motion control of a PUMA 560 robot. The simulation results and the performance of both off-line and online parameter identification are presented