Recent results in the development of optimal design of direct model reference adaptive controllers are summarized. Two methods have been developed: one employs an analytic averaging technique for solving a constrained nonlinear optimization problem yielding a close-form solution; the other uses a numerical optimization approach with high-level learning capability. These two approaches are outlined. A mathematical model of a flexible structure experiment facility was employed for testing the two design approaches. Numerical results are discussed and comparative analysis is performed to show the merit of these methods.
展开▼
