Dynamic Control of Three-Phase Dual-Active-Bridge DC-DC Converters Based on Sliding-Mode-Learning Adaline

Dynamic performance is an important index for three-phase dual active bridge (3p-DAB) DC-DC converters under the conditions of source voltage variance and load resistance step up/down. To enhance the dynamic response of output voltage tracking for 3p-DAB, a novel nonlinear controller, which is combined with the proportional plus derivative (PD) controller and the adaptive linear neuron (Adaline), is proposed in this article. Firstly, a sliding-mode-learning strategy is presented for the Adaline to improve the robustness of adaptive learning. Meanwhile, the output of the Adaline is added to the output of PD controller for ensuring an outstanding dynamic response of output voltage with less overshoot. Furthermore, a comparison with the PI controller based on phase shift modulation is implemented by simulation tool to confirm the theoretical analysis and validity of the proposed control scheme. Finally, experimental comparison results on a laboratory prototype are given to further validate the comparison results and the practical effectiveness of the designed nonlinear controller.