电压
电气工程
晶体管
整流器(神经网络)
电池(电)
电动汽车
氮化镓
功率(物理)
恒流
拓扑(电路)
材料科学
工程类
计算机科学
物理
图层(电子)
复合材料
循环神经网络
机器学习
随机神经网络
量子力学
人工神经网络
作者
Satyaki Mukherjee,Juan M. Ruiz,Peter Barbosa
标识
DOI:10.1109/tpel.2022.3217092
摘要
This article presents a single stage isolated bidirectional DC–DC converter comprising of LCL-T resonant network for universal electric vehicle (EV) charging. Fixed frequency operation along with phase-shift control enables easier design of passive components, while achieving high efficiency across wide output voltage range. Furthermore, to increase the switching frequency and consequently the power density, gallium nitride (GaN) switches are explored in the proposed solution. In order to facilitate use of commercially available 650 V GaN transistors, a multi-level inverter and a reconfigurable rectifier architecture is utilized along with LCL-T resonant network, accommodating an input voltage of 800 V, generated from an universal three phase ac input supply, and an output voltage ranging from 150 V to 950 V catering to large variation of battery voltages from different vehicle manufacturers. The constant input voltage-to-output current gain property of LCL-T network, along with an efficient phase-shift modulation proposed in this paper enable soft switching of all transistors with minimum circulating currents over constant current (CC), constant power (CP), and constant voltage (CV) modes of battery charging. Experimental results are provided for a single-phase DC–DC converter prototype utilizing 650 V GaN transistors operating from an 800 V DC bus and providing a very wide output voltage from 150 V to 950 V at 6.6 kW maximum power at a constant switching frequency of 500 kHz. The prototype achieves two peak efficiency points of 98.2% and maintains $>$ 97% efficiency across the entire output voltage range at a power density of 120 W/in 3 (7.3 kW/L).
科研通智能强力驱动
Strongly Powered by AbleSci AI