Extremum control for optimal operating point determination and set point optimization via sliding modes

In the design of a control system, it is often desirable to operate at the peak of an appropriate performance surface which characterizes the performance of the closed-loop operation. However, in many cases, only limited information might be available on the plant and the desired performance criterion which makes a priori determination of such an optimal operation mode difficult in the first place. The online identification of an optimal operating point and the development of a controller structure which enables the system to robustly operate at such a point constitute a remarkable research problem with this motivation. In this paper, a two-time scale sliding mode optimization method is studied for this purpose. The adopted scheme assumes a regulative controller which produces an equilibria for the closed-loop system parametrized by a free control parameter and employs a sliding mode optimization method to adapt this parameter in a slower time scale to increase the performance of the overall system. A simulation study is summarized to illustrate the approach. [S0022-0443(00)01004-2].