State and parameter estimation for dynamic systems with colored noise: an accuracy approach based on the minimum discrepancy measure

Abstract

An autoregressive moving average (ARMA) order-determination criterion based on a minimum discrepancy measure (MDM) is proposed to model colored measurement noise. Based on this criterion, an algorithm is presented to perform the state and parameter estimation. The criterion is consistent in the sense that the MDM decision rule will choose the correct order as the number of observations tends to infinity. Experimental results are included to illustrate the effectiveness of this technique. >