Subspace Identification of Closed-Loop Systems

Abstract

We briefly review a stochastic realization theory based subspace method for identifying closed loop systems [1, 2]. Using the preliminary orthogonal decomposition we show that, under the assumption that the exogenous input is persistently exciting (PE), the identification of closed loop systems is divided into two subproblems: the deterministic identification of the plant and controller, and the stochastic identification of the noise filter.