Eigen Value Computation Based Approach for Reduced Order Modeling for Interval Systems

Abstract

A new method of model order reduction technique for interval system is proposed here. The reduced order denominator coefficients are determined by combined benefits of reciprocal transformation, principal pseudo break frequency and a new method of eigen spectrum analysis. Then the reduced order numerator coefficients are obtained using polynomial coefficient matching method. Optimization method can also be used for determining the coefficients of reduced order numerator. In this paper Luus -- Jaakola optimization technique is tried. This method guaranteed a stable reduced order model for the stable higher order system along with performance matching. The illustrated examples using MATLAB simulation show the effectiveness of the proposed method.