Identification of Time-Delay Systems using the Subspace Method

Background/Objectives: A new algorithm for identification of LTI delayed systems by using the subspace method is proposed. The idea is based on removing the unknown triangular Toeplitz matrix by using the orthogonal of input Hankel matrix to obtain the observability matrix. Methods: The method is used for both deterministic and stochastic representations. Finding: The main feature of the present work over similar possible algorithms is that with a relatively low number of iterations we get the exact results. Application/Improvements: Two examples are shown to illustrate the validity of the method.