电气工程
整流器(神经网络)
电流(流体)
电压倍增器
计算机科学
电压
工程类
电压调节器
随机神经网络
机器学习
循环神经网络
人工神经网络
跌落电压
作者
Huu-Phuc Kieu,Dinh Bao-Hung Nguyen,Donghyuk Lee,Sewan Choi
标识
DOI:10.1109/apec48139.2024.10509124
摘要
This paper proposes a new magnetic integration approach that combines all the magnetic components, including the leakage inductor, transformer, and filter inductors in the current doubler circuit. The core structure proposed in this study is applicable to different topologies that incorporate the current doubler circuit. By adopting this all-in-one core structure, the assembly of the magnetic components is simplified, leading to a reduction in costs. The proposed core structure not only reduces the costs but also increases ZVS energy and reduces the output current ripple. Furthermore, the proposed structure also reduces the total magnetic volume by 15.7% and the total loss by 8.5% compared to the non-integrated. In order to verify the proposed magnetic integration structure, a 3kW auxiliary power module with the all-in-one magnetic structure has been built. The prototype achieves 8.4kW/l of power density and 95.5% of peak efficiency.
科研通智能强力驱动
Strongly Powered by AbleSci AI