A System Model for the Effect of Self-Balancing in Multiactive Bridge Converters

This article analyzes the effects of self-balancing in multiactive bridge converters. Therefore, it proposes a time-domain approach to model the dynamics of multiwinding transformer-based multilevel dc/dc converter, including self-balancing. In previous research, multiwinding transformers have been modeled using average-based power transfer models. These models, while simple, have no direct intuitive interpretation. Furthermore, they only include power transfer resulting from external phase shift between models, not taking the internal balancing of the shared magnetic core into account. In this article, a dc/dc converter model that includes the transformer as $M$-port network with magnetization and stray inductance and conduction losses is proposed. The model is verified using experiments showing both accurate predictions of both, circulating currents within each module and load currents affecting multiple modules. Furthermore, the effect of self-balancing is shown and how the model is capable to predict changes in the output voltage due to external loads. The model is also used to analyze the robustness of self-balancing to design parameter uncertainties.