A generalized fock space model for nonlinear systems driven by random inputs

A general nonparametric model for a non-linear dynamical system driven by a second-order random process is proposed. The model is based on the best approximation of the nonlinear system structure in a generalized Fock space as well as on the mean square estimation of the output given the input-output statistics. The model is realized by a cascade of a linear dynamical system L1, a nonlinear zero-memory system N, and a linear dynamical system L2.