Abstract: A controller consists of feedforward and feedback compensator. The term of feedforward is used to reduce the tracking error in the period of acceleration and deceleration. Feedback term is applied to improve the robustness, to compensate for the modeling error, and thus to guarantee the convergence of the tracking error. However in practice, the system is uncertainty due to the modeling errors, therefore an accurate feedforward coefficient cannot be obtained with standard method on the basis of the nominal model. Moreover, this MIMO(Multi-input multi-output) system cannot be decoupled and the convergence of the tracking errors is no longer guaranteed. In order to improve the servo performance of the planar motors with uncertainties, a feedforward identified method and nonlinear composite feedback controller are proposed, the convergence and stability of the control law are guaranteed by Lyapunov stability theory. In feedforward design, the feedforward coefficient can be identified by PD control experiment. For the nonlinear feedback controller, the robust term is designed to compensate for the modeling errors, therefore the coupling effects between multi-degrees of freedom are reduced. An amplitude-based variable-gain function is applied in the nonlinear term, thus the control action benefits from this operation by suppressing the low-frequent contribution and no extra control effect as to avoid the high-frequent noise. The trajectory tracking experimental results show that the proposed controller is an attractive approach to improve the servo performance of the system with uncertainties.

Key words: Modeling errors, Nonlinear composite controller, Planar motor, Trajectory tracking