On the identification of autoregressive moving average models

One of the problems encountered in the Box-Jenkins-type of adaptive control design is identification of noise that may follow an ARMA(p, q) model. In this article a new method based on the theory of Padé approximation has been introduced for the identification of ARMA models. In this method the estimates obtained by fitting an infinite AR model (or infinite MA model) are used to construct the so called C-table, which clearly depicts the order of an ARMA(p, q) model. Another advantage of this method is that we can simultaneously get estimates of the parameters of the identified model. Simulation results and results obtained from applying this method to real data sets substantiate our claims regarding this method.