Multiobjective design optimization of transformers for battery cell balancing converters considering bidirectional power flow

Abstract

Owing to an increase in the demand for bidirectional applications such as battery energy storage systems (BESS), isolated bidirectional converters have become more popular. However, conventional transformer design methods such as the area product and core geometrical coefficient methods, consider only one operating point. Thus, these do not always guarantee high efficiency if there are any changes in the operation point of the converter. Accordingly, this paper proposes a multiobjective optimized transformer design algorithm that considers the overall energy loss of the bidirectional operation of the converter. The proposed algorithm adopts a nondominated sorting genetic algorithm-II (NSGA-II) effective core cross-sectional area and turn ratios as main variables. A 20W prototype converter with a transformer for 20-series lithium-ion cell balancing purposes has been built for verification. The results show that the proposed algorithm dissipates lower power loss during the charge and discharge mode of operations with a smaller volume than the conventional method.