In this article, a novel Hammerstein structure is proposed for nonlinear networked systems based on an interval type-2 Takagi–Sugeno–Kang (IT2TSK) fuzzy system. The proposed approach is designed based on the general Hammerstein form, where an autoregressive moving average and an IT2TSK structures are designed as the linear energetic and nonlinear static components, respectively. The consequents of the nonlinear subsystem are characterized by a TSK-type system while the antecedents are characterized by interval type-2 fuzzy sets. The structure of the nonlinear subsystem is learned online based on the type-2 fuzzy clustering. Two new updating algorithms are introduced based on the Lyapunov theorem for learning the proposed model parameters and adaptive learning rates to assure the model stability and the parameter fast convergence. To illustrate the proposed model robustness, a test is performed under networked environment with time-varying delay and packet losses. The simulation results prove a higher performance for the proposed model than that of compared models.
Read full abstract