Decentralized robust control of a hydraulic driven parallel manipulator based on disturbance estimation

Since the parallel robot manipulator is a nonlinear, coupling and time-varying dynamic system, it is difficult to synthesize a controller based on the dynamic model by using a centralized control method. Considering each driving unit as a subsystem and treating the coupling interconnection terms between the joints as a time-varying external disturbance of the driving unit, this paper presents a decentralized robust controller with extended state observer (ESO) in joint-space to implement the high-precision trajectory tracking of the robot manipulator. By constructing two extended state observers, not only matched uncertainties but also unmatched uncertainties can be estimated and then compensated in a feed-forward way. Consequently, high-gain feedback is avoided. In addition, the feedback robust laws ensure the global robustness of the proposed controller. Furthermore, the stability of the system is verified by constructing the Lyapunov function. The simulation results show that the proposed controller can estimate both matched and unmatched uncertainties of the driving unit accurately, and the system possesses high trajectory tracking performance and strong robustness.