Subspace identification of Hammerstein systems with nonparametric input backlash and switch nonlinearities

Abstract

In this paper, a subspace identification algorithm for a class of Hammerstein systems is developed. We consider dynamical systems subject to input backlash or switch nonlinearities. The idea is to use a specific input signal allowing the estimation of the nonlinear part and the estimation of a state space model for the linear part. The identification algorithm is a subspace type algorithm. A simulation example is given to illustrate the performances of the present method.