A Hybrid Optimization Algorithm for the Design of DC Resistant Medium-Frequency Transformers

The Medium-Frequency Transformer (MFT) is a key - but often limiting - element in the design of isolated DC-DC converters. The design process of an MFT is computationally complex and converter asymmetries can lead to core saturation, limiting DC- DC converter operation. This paper presents a method to adjust the transformer main inductance by utilizing an air-gapped core to shift the core saturation point and increase the MFT's ability to withstand DC offsets. The adjustment of the main inductance is achieved with only minor compromises in mass and efficiency. Furthermore, this paper introduces a hybrid multi-objective optimization algorithm based on genetic algorithm and particle swarm optimization to solve the multi-dimensional MFT design problem. It is demonstrated that the hybrid algorithm requires more iterations to outperform the individual algorithms but allows the design of DC resistant MFTs over a wide range in the objective space with a tight solution distribution.