Optimization of the pulse transformer using circuit-field model

The paper presents the new strategy of the optimization of pulse transformer (PT). In order to reduce calculation time the optimization problem has been decomposed into two stages. In the first stage, to determine functional parameters of PT the circuit model is used. The goal of circuit calculations is to limit the space of design variables that meets formulated requirements. The genetic algorithm has been applied for this task. In order to include constraints, the penalty function has been engaged. The transformer dimensions obtained in the first stage of calculations are used as initial values in the second stage of design process. In the second stage the field model of PT is employed. Obtained results prove that presented approach allows for fast optimization of the PT design.