Model Order Reduction Of Discrete Time Systems Using K-Means Algorithm
A new method is proposed in this paper for reducing the order of linear discrete time system. The proposed
method uses fuzzy K-means algorithm for obtaining the denominator polynomial and matching of coefficient ratio for
formulating numerator polynomial. Bilinear transformation is used for the analysis of the system in w-plane. The proposed
method results a reduced order model which retains the stability of the original system and typical characteristics of the
original system and avoids calculation of time moments and formulator of tables which is necessary in some of the existing
Keywords— Model Order Reduction, Discrete Systems, Pole Clustering, K-means.