Recursive closed-loop subspace identification based on predictor Markov parameters

Abstract

Recursive subspace identification methods have been an active area of research in recent years. In this paper, a recursive version of a closed-loop subspace method is proposed for on-line system modeling. The algorithm is based on the predictor Markov parameters and is prone to a very simple and computationally attractive implementation. A numerical example shows the effectiveness of the proposed algorithm on the problem of closed-loop recursive identification.