Identification of Polynomial Wiener Systems via Volterra-Laguerre Series with Model Mismatch

Abstract This paper considers the effects due to model mismatch in system identification of discrete polynomial Wiener systems via Volterra-Laguerre series. A special case of third-order polynomial nonlinearities and Gaussian input signals is treated in detail though the approach appears to be valid for a more general problem formulation. Expressions for the bias in the estimates of the model coeficients due to a linear block model mismatch are derived and an iterative algorithm for reducing the bias value is proposed. The analytical results are illustrated by numerical experiments confirming the eficacy of the bias reduction algorithm.