Stable and optimal adaptive fuzzy control of complex systems using fuzzy dynamic model

A design method is presented for the stabilization of multivariable complex nonlinear systems which can be represented by fuzzy dynamic model. Firstly, an adaptive fuzzy control law is addressed. Then the closed-loop fuzzy system is robustly stabilized under some bounded perturbations and disturbances by adding a supervising controller. Based on the variable structure control theory of uncertain systems, a sufficient condition for the stability of the closed-loop fuzzy system is also derived using the concept of connective stability in decentralized control of large-scale systems. Under this condition, an optimal fuzzy controller is designed with genetic algorithms. Two examples are given to illustrate the application of the method.