Set-Membership identification of Wiener models with non-invertible nonlinearity

Authors:	Cerone Vito, Politecnico di Torino, Italy Regruto Diego, Politecnico di Torino, Italy
Topic:	1.1 Modelling, Identification & Signal Processing
Session:	Set-Membership Techniques for Identification
Keywords:	identification, nonlinear models, parameter estimation, uncertainty, bounded noise

Abstract

In this paper the identification of SISO Wiener models in presence of boundedoutput noise is considered. A three stage procedure based on the inner signalestimation, outlined in a previous contribution, has been proposed. Results andalgorithms for the computation of inner-signal bounds through the design of asuitable input sequence have been provided for the case of polynomial withoutinvertibility assumptions. A simulated example, showing the effectiveness ofthe proposed approach, has been presented.