Application of Artificial Intelligence in Optimization of Solid State Transformer Core for Modern Electric Vehicles Using Multi-Objective Genetic Algorithm

Abstract

Solid State Transformers (SST) are increasingly becoming a favored alternative to traditional low frequency transformers due to their small size and excellent efficiency, particularly in the field of Electric Vehicles (EV), which has seen rapid growth in recent years. In this research, we propose a multi-objective Artificial Intelligence-based high-frequency transformer (HFT) design optimization for solid state transformer (SST) applications. As the key component of the SST, the optimization of the HFT design parameters is crucial to achieving high efficiency and power density, irrespective of its topology. The HFT is designed using a multi-objective non-dominated sorting optimization technique that minimizes core volume (maximizing power density), total transformer losses, and the overall cost based on a set of numerous Pareto-Optimal Solutions (POS). In this study, an 85 kHz, 10kW HFT with several high permeability core materials are investigated, and the POS are investigated. The findings show how the various design variables affect the goal functions. The results further demonstrate that by carefully selecting design variables using the proposed method, the size, efficiency, and cost of the HFT may be efficiently maximized. A significant number of pareto-optimal solutions show that an efficiency of more than 97.5 % can be achieved in the HFT design for SST. The hardware implementation of the optimized HFT design is currently underway and will be presented in future publication.