Identification of Wiener nonlinear systems using the key-term separation principle and the filtering approach

A Wiener finite impulse response moving average (FIR-MA) model is rewritten as a standard regression controlled moving average form by applying the key-term separation principle. A filtering based recursive least squares estimation algorithm and a maximum likelihood recursive least squares estimation algorithm are developed for the parameter identification of Wiener FIR-MA model. Both of the proposed algorithms could reduce the influence of colored noise in parameters estimation through the data filtering approach; further improve the accuracy of parameters estimation. The simulation results demonstrate the efficiency of two proposed algorithms.