Fuzzy iterative learning identification algorithms of time-varying nonlinear systems

Two fuzzy iterative learning identification algorithms are presented, in this paper, for modeling and identification of continuous time-varying nonlinear systems. The algorithms are used to adjust the parameters involved in the fuzzy systems, by means of the iterative learning manner, where the system undertaken runs over a finite interval repeatedly. The compensation is introduced in the learning mechanism to eliminate the influence of the approximation error. With the use of time-varying fuzzy systems for the identification of time-varying nonlinear systems, the less number of fuzzy rules is expected, which helps to reduce the online computation of the identification process. In this paper, Lyapunov approach is used for the identifier design and the convergence analysis. The identification error is ensured to converge to zero over the entire interval after a number of iterations, and all the parameter estimates are guaranteed to be bounded.