Model order reduction of continuous interval systems via time moment matching and pole clustering approaches

Abstract

In this manuscript, authors propose an algorithm for model order reduction of interval systems. Algorithm utilizes pole clustering and time-moment matching approaches for calculating denominator and numerator of model respectively. In pole clustering approach, poles of the high order system are considered for denominator calculation. According to order of the model, clusters of the poles are framed. Cluster centre of every cluster is determined using inverse distance criterion. Using these cluster centres, model denominator is deduced. Numerator is obtained by equating time moments of system and model. Proposed algorithm is applied on sixth order system and results are compared with other existing techniques which shows that proposed algorithm is superior to other techniques.