Multivariable Predictive Control Based on the T-S Fuzzy Model

Abstract

There are many complex industrial processes, such as the load control system of a power plant, that have nonlinear dynamics with time-varying parameters and with large time-delays. It is usually very difficult to design a satisfactory control system for such processes []. The adaptive control of nonlinear systems is one of the most often applied methods. In most cases, this approach is to transform nonlinear system dynamics into an appropriate linear model around an operating point, so that conventional linear control techniques can be applied []. A key assumption in these studies is that the system nonlinearities are known a priori and they are linearizable. Such an assumption limits the applications of the theory because real systems always contain uncertain disturbance and unmodeled dynamics. The design of a highly accurate modeling method for nonlinear systems and a nonlinear model-based adaptive control methods helps to deal with these limitations.