Identification of Fractional Model by Least Squares Method and Instrumental Variable

Abstract

This paper deals with fractional model identification using leastsquares (LS) method and instrumental variable (IV) in a noisy output context. A new identification method, which extends LS techniques to fractional system to identify not only the parameters but also the unknown order, is presented. In order to eliminate the bias of identification results, IV method is chosen which permits unbiased parameter estimation. Monte Carlo simulation analyses are used to demonstrate the validity and the performance of the proposed fractional order system identification method.