Optimization of LTI Systems Using Modified Clustering Algorithm.

A novel order diminution method for linear time invariant continuous systems is proposed in this paper. The reduced denominator polynomial is obtained by modified clustering algorithm and the reduced numerator polynomial coefficients are generated by an evolutionary algorithm as referred in this communication. Several numerical examples taken from the literature have been solved for the validation of the proposed method. The reduced system obtained by proposed method guarantees the stability if the original system is stable and also preserves all characteristics of given higher order system in the reduced one. The presented method is to be continued to multivariable systems and illustrates with numerical examples for proving the efficiency of the proposed method. Furthermore, the proposed method is also to be extended for order reduction of discrete time systems. [ABSTRACT FROM AUTHOR]