A Unified Modeling Approach for Design and Performance Improvement of Triple Active Bridge Converter
Abstract
Triple Active Bridge (TAB) converter is a multi-port DC-DC converter. This converter is an extension of the popular Dual Active Bridge converter. It features desirable traits of the DAB converter, such as high power density, galvanic isolation, and bi-directional power flow between any of the ports. As in other multi-port converters, redundant power conversion is minimized through component sharing among the ports in a TAB converter. All the switches in a TAB converter can undergo soft-switching over a wide range of operating points, reducing switching losses and the size of auxiliary components. The multiple degrees of freedom in modulating a TAB converter offer several design and operational flexibilities.
However, this converter has yet to come into the limelight despite these advantages. One of the reasons is the lack of a unified analytical framework for the design and operation of this converter. The existing models for the TAB converter are limited in scope and cannot be easily used for the design and operational optimization of the converter. This work focuses on developing simple, unified models for analyzing the TAB converter.
Firstly, the popular Fundamental Harmonic Approximated (FHA) large-signal and small- signal models are evaluated to understand their limitations. Improved large-signal and small-signal Generalised Harmonic Models (GHM) are developed by incorporating the impact of higher-order harmonics. While the GHM is shown to be superior for small-signal analysis of the converter and the design of a closed-loop control system, it is not suitable to analyze the soft-switching bounds of the TAB converter. To overcome the limitations of GHM, a Unified Model that incorporates the impact of the magnetising inductance of the three-winding transformer is proposed. The Unified Model can accurately predict the AC port currents at the switching instants and is used to study the soft-switching bounds of the TAB converter. The GHM and Unified Model are validated through extensive switching circuit simulations and experimental results from a 1 kW hardware prototype developed in the laboratory.
Further, a new design algorithm for the TAB converter is proposed. The proposed algorithm leverages the FHA model’s simplicity and the Unified Model’s accuracy. Finally, a new modulation scheme based on Penta Phase Shift with five degrees of freedom is proposed to achieve soft-switching across the operational range of the TAB converter.