Reduced-order modeling of commensurate fractional-order systems

Abstract

This paper proposes a new method of finding reduced-order model of a commensurate fractional-order (FO) system. The FO system is first converted to integer-order (IO) and then a mixed method of model-order reduction (MOR) scheme is applied. In this scheme, the denominator is obtained using time-moment matching method and the numerator is determined using the recently developed big bang-big crunch (BB-BC) optimization algorithm. This formulated reduced-order model is reconverted into its commensurate form of FO model. Three examples are presented to validate and illustrate the efficacy of the proposed approach.