Design Methodology for Symmetric CLLC Resonant DC Transformer Considering Voltage Conversion Ratio, System Stability, and Efficiency

With the popularity of dc microgrid, the symmetric CLLC resonant dual active bridge converter usually works as a dc transformer (DCT) when it is applied in dc microgrid. It has received increasing attention as it provides galvanic isolation and also enjoys high power density. In such application, the open-loop control with 50% duty ratio is commonly adopted for the CLLC resonant DCT (CLLC-DCT) for simple control scheme and high efficiency. However, this will bring about difficulties for the design of CLLC-DCT. First, the fluctuating values of the practical inductors or capacitors because of temperature or power variation may lead to deviated voltage conversion ratio (VCR) of CLLC-DCT. Second, chances are that system instability will happen when CLLC-DCT and a constant power load connect in a cascaded structure because of impedance interactions, which will also be affected when parameters fluctuate. Third, high efficiency of the DCT is also expected even when the inductance or capacitance fluctuate. With all the above concerns, the five-stage design methodology for the CLLC-DCT converter is put forward to meet design objectives regarding VCR, system stability, and power efficiency simultaneously. And the effects of fluctuating inductance and capacitance have been fully taken into account by the assistance of particle swarm optimization. Besides, a 1-kW hardware prototype has experimentally validated that the proposed design methodology is feasible.

Automated summary

With the increasing popularity of dc microgrid, the symmetric CLLC resonant dual active bridge converter is commonly used as a dc transformer (DCT) in such applications. However, challenges arise in the design of CLLC-DCT, including deviation in voltage conversion ratio (VCR) due to fluctuating values of inductors or capacitors, system instability when connected to a constant power load, and the need for high efficiency even with parameter fluctuations. A five-stage design methodology has been proposed to address these concerns, using particle swarm optimization for optimization and validation through a 1-kW hardware prototype.