Modeling and identification of a fractional-order discrete-time SISO Laguerre-Wiener system

This paper presents a new implementable strategy for modeling and identification of a fractional-order discrete-time nonlinear block-oriented SISO Wiener system. The concept of modeling of a linear dynamics by means of orthonormal basis functions (OBF) is employed to separate linear and nonlinear submodels, which enables a linear regression formulation of the parameter estimation problem. Finally, discrete-time Laguerre filters are uniquely embedded in modeling of the fractional-order dynamics, eliminating the disastrous bilinearity issue. Simulation experiments show a very good identification performance for a fractional-order Laguerre-based Wiener model, both in terms of low prediction errors and accurate reconstruction of the actual system characteristics.