A Renovated Pole Clustering Technique for Model Order Reduction

Abstract

A renovated pole clustering technique is proposed in this paper to reduce large scale linear dynamic systems into lower order linear dynamic systems. The denominator polynomial order is decreased by using renovated pole clustering technique and numerator coefficients are calculated by using improved pade approximations technique. The reduction algorithm is completely computational. The reduced system retained the stability characteristic of the original higher order system. The Comparison is performed by simulated graph and performance index like integral square error (ISE) which justify its excellence.