About nonparametric algorithms of nonlinear dynamical processes identification

The authors consider a problem of nonlinear dynamical systems of Wiener-Hammerstein type identification, when linear dynamical part of the system is in conditions of nonparametric uncertainty. The type of nonlinearity is assumed to be unknown or known completing the parameters. The algorithm for creation of adequate, as root mean square criterion, for models of systems. The proposed method of dynamic objects modeling is on nonparametric estimation of linear and nonlinear sub-systems.