Recursive identification for Hammerstein and Wiener systems with piece-wise linear memoryless block

Author:	Chen Han-Fu, Chinese Academy of Sciences, China
Topic:	1.1 Modelling, Identification & Signal Processing
Session:	Nonlinear System Identification - Kernel Methods
Keywords:	Wiener system, Hammerstein system, identification, strong consistency, kernel function

Abstract

The paper deals with identification of Hammerstein and Wienersystems with nonlinearity being a discontinuous piece-wise linearfunction. Recursive estimation algorithms are given to estimatesix unknown parameters contained in the nonlinearity and allcoefficients of the linear subsystem. The estimates converge to the corresponding true values with probability one.Numerical examples are given to verify the theoreticalassertions. Copyright © 2005 IFAC