Singular Perturbation based Frequency-Weighted Model Reduction of Discrete-Time Systems

Abstract

In this paper, we present a new frequency weighted model reduction technique for discrete-time systems using balanced singular perturbation approximation with the help of partial fraction expansion. The proposed method ensures stability of the reduced-order models in double-sided weighting condition. The basic idea of the proposed technique is to merge the unweighted balanced technique and the partial-fraction based frequency-weighted technique. A $6^{\mathrm {t}\mathrm {h}}$ order stable low pass digital elliptic filter is considered to show the results of the proposed method.