Refined instrumental variable methods for identification of LPV output-error and Box-Jenkins models
Résumé
Identification of Linear Parameter-Varying (LPV) systems in an Input-Output (IO) setting is investigated, focusing on the case when the noise part of the data generating system is an additive colored noise. In the Box-Jenkins (BJ) and Output-Error (OE) cases, it is shown that the currently available linear regression and Instrumental Variable (IV) methods from the literature are not optimal in terms of bias and variance of the estimates. To overcome the underlying problems, a Refined Instrumental Variable (RIV) method is introduced.