On LPV System Identification Algorithms for Input-Output Model Structures and their Relation to LTI System Identification

摘要

Two new identification algorithms for linear parameter-varying (LPV) input-output (IO) systems are introduced in this paper based on the ARMAX and BJ model structure, respectively. Both algorithms are applicable for the MIMO case and thus fill the gaps of existing LPV IO system identification theory. Moreover, the relation between identification of LPV and LTI models is pointed out and exploited for the new LPV ARMAX algorithm. Efficacy of the two new algorithms is shown in simulation examples.