Application of adaptable functional series vector time-dependent autoregressive model for extraction of real modal parameters for identification of time-varying systems

Abstract This paper presents a method for extraction of real modal parameters for identification of time-varying systems using adaptable functional series vector time-dependent autoregressive (AFS-VTAR) model. In the practical application of AFS-VTAR model, both real modal parameters and computational ones may mix together in the identified results. In order to extract the real ones from computational ones, a new method based on the eigen-decomposition of the state matrix constructed in AFS-VTAR is proposed. The efficiency of the proposed method is successfully demonstrated first through numerical simulations applied to a time-varying 3-degree-of-freedom system and then experimentally by applying it to a flexible robot in motion.