Multiobjective optimal design of transformers for isolated switch mode power supplies

Abstract

This paper deals with the multiobjective optimal design of medium frequency (MF) transformers for isolated switch mode power supplies (SMPS). The design problem requires minimizing the weight and the loss of the transformer while ensuring the satisfaction of a number of constraints. These constraints include appropriate limits on efficiency, temperature rise and winding dimensions. The optimization variables are, as for them, the operating frequency, the maximum flux density, the current density in windings as well as the core dimensions. If the first three variables are commonly used as design variables, it is rarer to treat the core dimensions as design variables. Note that the MF transformer model used in the design procedure takes full account of the current and voltage waveforms. In this contribution, a design procedure of MF transformers based on multiobjective Genetic Algorithms (GAs) is proposed and illustrated with the design of a transformer for a full-bridge converter.