Subspace identification forwiener systems with general nonlinearity

In this paper, the subspace methods for Wiener systems with state space model are considered, in which a general condition for the static nonlinear maps is proposed. By means of differential analysis and Singular Vector Decomposition, the subspace of extended controllability matrix can be obtained, and then All parameter matrices in linear subsystems and nonlinear function can be estimated. In mild conditions, we show that all estimates are consistent in some sense. A simulated example is provided to verify the method proposed in this paper.