Auxiliary model maximum likelihood least squares-based iterative algorithm for multivariable autoregressive output-error autoregressive moving average systems

Qian Zhang, Huihui Wang, Ximei Liu

2024Proceedings of the Institution of Mechanical Engineers Part I Journal of Systems and Control Engineering24 citationsDOI

Abstract

Identification of multivariable systems is of great significance to control systems. This paper focuses on the parameter identification problems for multivariable autoregressive output-error autoregressive moving average (M-AROEARMA) systems. On the basis of the decomposition strategy, the M-AROEARMA model is de- composed into multiple subsystem models. By means of the auxiliary model idea, the auxiliary model least squares-based iterative algorithm is derived. For the purpose of achieving highly accurate parameter identification performance under colored noises interference, an auxiliary model maximum likelihood least squares-based iterative algorithm is proposed by utilizing the maximum likelihood principle. The numerical simulation example demonstrates the effectiveness of the proposed algorithms.

Topics & Concepts

Autoregressive modelSTAR modelSETARMultivariable calculusNonlinear autoregressive exogenous modelMathematicsAutoregressive–moving-average modelLeast-squares function approximationAlgorithmApplied mathematicsMaximum likelihoodComputer scienceStatisticsAutoregressive integrated moving averageTime seriesEngineeringEstimatorControl engineeringControl Systems and IdentificationStructural Health Monitoring TechniquesAdvanced Adaptive Filtering Techniques