Optimal Learning Tracking Control for Nonlinear Systems

摘要

In this paper, an improved-type learning control mechanism is proposed to design an optimal tracker for a nonlinear system. Even without the detailed information on the system, the proposed mechanism also yields the optimal learning tracker. In order to achieve the output tracking design, the optimal linearized models of the nonlinear system at operating states are obtained by the optimal linearization approach, so that the well-designed learning mechanism for linear systems can be extended to the nonlinear case. How the eigenspectrums of linearized models significantly affect the effectiveness of the optimal learning control mechanism has been investigated in this paper, and the pole placement method is used to improve the drawback of learning control. Illustrative examples are presented to demonstrate the effectiveness of the proposed methodology.