Continuous Time State Space Model Identification Using Closed-Loop Data

This paper focuses on identifying a continuous time state space model for a system operating in closed-loop, using a subspace method based on error-in-variables (EIV) models. The proposed approach in this paper extends the existing methods in the discrete-time systems to continuous-time systems where the Laguerre filters are used in the identification procedure. Furthermore, to meet the requirement for continuous time model and to remain filter causality, the choice of instrumental variable is based on the future horizon variables. Monte-Carlo simulation results are presented to verify the consistency of the estimated models.