An Adaptive Type-2 Fuzzy Neural Controller for Nonlinear Uncertain Systems

Abstract

This article proposes a new control scheme using type-2 fuzzy neural network (type-2 FNN) and adaptive filter for controlling nonlinear uncertain systems. This type-2 FNN model combines the advantages of type-2 fuzzy logic systems and neural networks. The type-2 FNN system has the ability of universal approximation, that is, identification of nonlinear dynamic systems. We herein adopt it to develop a novel control scheme for nonlinear uncertain systems. The proposed control scheme consists of a PD-type adaptive FNN controller and a pre-filter. The adaptive filter is used to provide better performance under transient response and to treat the problem of disturbance attenuation. The tuning parameters for the filter and the type-2 FNN controller will change according to the learning algorithm. By the Lyapunov stability theorem, the convergence of parameters is given in order to guarantee the stability of nonlinear uncertain systems. The effectiveness of the proposed controller is demonstrated by simulated results.