Design of a Wound Core Pulse Transformer Using Multiobjective Optimization Method

Abstract

In design of a pulse transformer, there are several conflicting objective functions that should be optimized. However, due to the high complexity, the conventional design methods cannot consider all of them simultaneously. In this paper, multiobjective particle swarm optimization is used to design a pulse transformer. This method is based on Pareto optimality. The output consists of a set of optimal solutions that the best design is selected from it. To use this optimization method, a model is developed for a pulse transformer with toroidal wound core. The model considers magnetic, electric, and thermal characteristic of a transformer. Its simple calculations make it proper for using in an optimization process. A transformer is designed and implemented for minimum rise time, temperature rise, and volume using the presented method. This transformer is compared with a transformer that has been designed by conventional methods. Measurements and comparisons show that the results are more acceptable than the results of convectional designs besides simplicity and accuracy of the method.