| I. System identification. The ARMA (autoregressive moving average) model plays an important role in system identification which is described as: yn=- i=1ai yn-i+j= 0bjx n-j+en ; The ARMA model is broadly used in many diverse fields ranging from signal processing, communications, biomedicine, to economics.; The true ARMA model order (p,q) is unknown; therefore, to circumvent this inherent limitation with the ARMA model predication, we have developed two novel algorithms to obtain model order.; II. Parameter estimation. Once we have obtained the accurate ARMA structure, the next step is to calculate the parameters. The conventional methods are generally biased. We have developed a new algorithm to overcome this shortcoming. |
