Two-level fuzzy control of large-scale systems and its application into PSS design

Abstract

This paper proposes a two-level suboptimal control using fuzzy prediction to control large-scale systems. A class of large-scale linear systems composed of interconnected subsystems is investigated. The overall control problem that is posed as a minimization of overall objective function, which is considered to be of quadratic form, is reduced to some optimization problems of lower order (sub) systems. The control input of each subsystem is composed of two signals. The first represents the local control signal (first level) and the second is the prediction signal (second level). In fact, the second signal is the prediction of interaction of other subsystems. It applies to each subsystem at every specified sample time (coordination sample times). The fuzzy logic theory is used for interaction prediction, where the prediction signal is constructed by a set of fuzzy sets with respect to state variables in an appropriate inference engine manner. The number of fuzzy sets and their interval deviations vary with time. Finally, the proposed method is applied to a three-area power system for designing a power system stabilizer (PSS)