Model order reduction of transformer linear section model using Simulated Annealing

This brief considers the order reduction of transformer linear section model using combined advantages of Simulated Annealing (SA) and Dominant pole retention technique. SA is a local search method, inspired by the process of physical annealing associated with solids. SA technique in association with dominant pole retention technique shows its effectiveness in obtaining low order model of transformer such that reduced order model approximates the original higher order transformer model and maintains the stability of the original system. The SA optimization method is based on minimization of the integral square error (ISE) between the transient responses of original higher order and the reduced order transformer linear section model. The zeros of reduced order model are obtained using SA by minimizing the ISE between high order and low order models of transformer, while poles are determined by dominant pole retention technique. The algorithm has been illustrated by one transformer model available in the literature. The step and frequency responses of both low and high order transformer models have also been compared along with the transient response parameters and a comparative study of ISE with the other existing techniques has also been given in the tabular form.