Recursive identification of Wiener systems with nonlinearities

Abstract

This paper deals with the identification of Wiener systems with hard nonlinearities. Using the switching function, a special form of nonlinearity representation is used in the Wiener system description. Two recursive identification algorithms are proposed, one is the recursive stochastic gradient identification algorithm and the other is the adaptive forgetting through multiple models identification algorithm. Eventually the performance between these two algorithms has been compared by experiment simulations.