Multi-Objective Optimization of Bi-directional On-Board Chargers Based on 650V GaN Power Transistors

High voltage GaN power transistors can significantly improve the key performance indices of a power electronic converter in terms of efficiency, reliability and power density. This brings a potentiality for widespread adoption of these semiconductors in e-mobility applications such as on-board chargers (OBCs) in electric vehicles. According to literature surveys, among all semiconductor technologies GaN-based high-electron-mobility transistors (HEMTs) devices are utilized to design compact, efficient and reliable OBCs. In this paper, a component-level (i.e., switch, capacitor and magnetics) optimization model is proposed based on Multi-objective Genetic Algorithm (MOGA) technique to assess the relationship between size (cm3), losses (W), and reliability (number of cycles) of the components used in OBCs. Besides, electrical circuit models establishing the relationship between these components for a dual-stage OBC architecture are presented. Finally, Pareto-front solutions for these components under the worst-case operation scenario are presented for the components used in the AC-DC stage of a bidirectional phase-modular three-phase OBC with a rated power of 11 kW.