Parameter Estimation for a Class of Multiple-Input Multiple-Output Systems Using the Maximum Likelihood Principle

Identification methods for multiple-input multiple-output systems depends on their structures and number of parameters. This paper aims at proposing new methods for estimating multiple-input multiple-output autoregressive output-error autoregressive moving average systems. A recursive generalized extended least squares method is presented for parameter estimation. In order to improve the estimation accuracy, the maximum likelihood recursive generalized extended least squares (ML-RGELS) method is proposed by using the maximum likelihood principle. The simulation results demonstrate that the proposed algorithms are all effective, and the ML-RGELS algorithm has the higher estimation accuracy.