Identifying the fractional-order systems with frequency responses: A maximum likelihood algorithm

System modeling plays an important role in radar servo system design with high precision. In this article a maximum likelihood identification algorithm of fractional-order systems with frequency responses is presented. The algorithm is deduced in detail and an equivalent criterion function is obtained from primary likelihood function. Both of the model coefficients and differentiation order estimation are considered in this paper, where two methods of estimating the fractional commensurate order are discussed. One is order scanning method and the other is simultaneous estimation of model coefficients and the differentiation order. Several simulation results are presented, together with the analysis of noise influences.