Multi-Physical Field Optimization Design of High-Frequency Transformers Based on the NSGA-II Algorithm

High-frequency transformers (HFTs), as a key device to support the efficient operation of smart grids, are indispensable in realizing core scenarios such as high percentage renewable energy access and flexible transmission. In this paper, a HFT design methodology is proposed based on multi-physics field coupling and multi-objective optimization. By introducing coupled analysis of the thermal, electromagnetic, and mechanical fields, a non-dominated sorting genetic algorithm (NSGA-II) is employed to perform multi-objective optimization of the transformer's power density, efficiency, and cost, aiming to achieve the optimal balance among these factors. A multi-physics field coupling model for the HFT is established, and the impacts of electromagnetic, thermal, and mechanical factors on transformer performance are analyzed. Then, a systematic design optimization is conducted using the optimization algorithm. The results show that the optimized transformer’s power density increase to 56.82×106W/m3, efficiency improve to 97.92%, and the cost remains well-controlled at a low level. Further analysis of a 1 MVA, 10 kHz high-frequency transformer shows that the long-term stability and reliability of the transformer can be improved by considering the coupling effects of multiple physical fields.