IDENTIFYING MIMO HAMMERSTEIN SYSTEMS IN THE CONTEXT OF SUBSPACE MODELIDENTIFICATION METHODS

In this paper, we outline the extension of the MOESP family of subspac e model identification schemes to the Hammerstein-type of nonlinear sy stem. Two types of identification problem are considered. The first ty pe assumes the (polynomial) structure of the static nonlinearity to be given and the task is to identify both the linear system dynamics and the unknown proportional constants in the parametrization of the stat ic nonlinearity. The second type addresses the identification of both the linear dynamic part and the static nonlinearity, where only limite d a priori information regarding the structure of the nonlinearity is available. The improved robustness properties of the algorithms develo ped for this second type of Hammerstein identification problem over ex isting correlation-based schemes is illustrated by a numerical example .