Advanced Modeling and Optimization of Multi-Active Bridge Converters

Theoretical models of MAB converters usually assume ideal components, but real-world systems suffer from various parasitic effects that skew power transfer predictions. These include MOSFET on-state resistances, AC winding losses in inductors/transformers, and the ESR of both DC blocking and DC link capacitors, all of which introduce measurable conduction losses and voltage drops. Additionally, since DC blocking capacitors have frequency-dependent impedance, they do not behave as ideal short circuits at all frequencies. Thus, there is an inadequate representation of converter non-idealities in traditional models of MAB converters resulting in inefficient power flow control due to simplistic modulation techniques.

 

Researchers at Arizona State University have developed an enhanced unified modeling framework for analyzing and modeling multi-active bridge (MAB) converters by including real-world non-idealities such as resistive losses, dead time effects, and transformer parasitics. This framework was designed to accurately predict power flow and current behavior as well as fine-tune control parameters to improve converter efficiency in real-time digital control systems. It enables accurate representation of both resonant and non-resonant transformer-isolated H-bridge-based converters while taking into account non-idealities that are common in circuits.

 

This unified modeling framework for multi-active bridge converters incorporates non-idealities and optimizes control to enhance efficiency and accuracy.

 

For more information about this opportunity, please see