Improved algorithm of model reduction of large-scale internal system

Recently, genetic algorithm for generating low order models for large-scale interval systems using simplified polynomial improved algorithm is proposed. New algorithm are developed for obtaining the numerator and denominator polynomials of the reduced order model which also retain the steady state response characteristics of the original high order interval system in its low order model, unlike existing methods. The method is computationally simple avoiding the necessity of formation of long Routh-type differentiation arrays and reciprocal transformations and always retains stability of the original system in the low-order models. Typical numerical illustrating examples are provided by using the proposed method.