Statistical Analysis of a Third-Order Cumulants Based Algorithm for Discrete-Time Errors-in-Variables Identification
Résumé
This paper deals with identification of dynamic discrete-time errors-in-variables systems. The statistical accuracy of a least squares estimator based on third-order cumulants is analyzed. In particular, the asymptotic covariance matrix of the estimated parameters is derived. The results are supported by numerical simulation studies.
Origine : Fichiers produits par l'(les) auteur(s)