Genetic Algorithm Assisted Parametric Design of Splitting Inductance in High Frequency GaN-based Dual Active Bridge Converter
Chang Wang,Tiberiu-Gabriel Zsurzsan,Zhe Zhang,Gabriel Tiberiu Zsurzsan
DOI: https://doi.org/10.1109/tie.2021.3102398
IF: 7.7
2021-01-01
IEEE Transactions on Industrial Electronics
Abstract:Splitting and placing interfacing inductance on both sides of the transformer has been proven to be an effective method, which extends the zero-voltage switching region for all the switching devices in the dual active bridge (DAB) converter. With the trend toward operating in higher frequency, achieving higher power density and higher efficiency, the converter model becomes more complex due to the non-negligible parasitic components that brings new challenges to DAB converter design. Traditional analytical methods have made it hard to imitate the proposed converter easily and precisely. Thus, artificial intelligence techniques are able to be utilized to assist the design process. When considering the converter system as a gray-box model, the metaheuristic algorithm can be implemented for the targeted design inside such a gray-box. In this article, a genetic algorithm (GA) is employed in the DAB converter parametric design with an explicit fitness desire to help in discovering the high frequency oscillation (HFO) problem. Consequently, the splitting inductance tuning method is proposed for eliminating the HFO problem and minimizing inductors' loss. The methodology of implementing GA into converter parametric design, and the proposed splitting inductance tuning method are introduced and verified with a 1 MHz gallium nitride high-electron-mobility transistor based DAB converter prototype. The comparitive experimental results prove the effectiveness of the splitting inductance tuning method and achieve 4% efficiency enhancement with 200 W power delivering.
automation & control systems,engineering, electrical & electronic,instruments & instrumentation