Recursive Extended Least-squares Method for Unity Feedback System

In this paper, the problem of unbiased recursive identification of a plant model in closed-loop operation with unit feedback is considered. A recursive extended least squares (RELS) algorithm for unity feedback system is presented by means of the model equivalence idea. The proposed algorithm is simple and efficient and can give the unbiased estimation of the plant model parameters in closed-loop system, further more, and can be on-line implemented. The advantages of the proposed RELS algorithm over the other approach are shown by simulation tests. Finally, this paper gives some relevant future works about RRLS method for closed-loop identification.