Model Predictive Control with Adaptive Frequency and Phase-shift Steps for Dual Active Bridge Converters

Dual Active Bridge (DAB) converters have gained significant popularity in isolated DC-DC conversion applications. However, their efficiency at light loads remains significantly low. Multiple control variables have been used to reduce the current stress of transformers, thereby improving power conversion efficiency of DAB converters, but at the cost of complexing the control algorithm design. To address this problem, this paper presents a simple Model Predictive Control (MPC), regulating only two control variables, i.e. the switching frequency and the phase shift, for high efficiency while maintaining good output voltage tracking. The proposed method discretizes these continuous control variables to create a finite control set, making the MPC easy to implement in digital processors. The optimal control variables for low current stress are derived analytically and incorporated into a meticulously designed cost function for multi-objective control. To trade-off between the dynamic and static performance, the adaptive step technique is applied to both control variables in the optimization process of MPC, driving the search in two dimensions of the control domain. Simulations in PLECS show that the proposed method achieves high conversion efficiency across a wide output power range, especially at light load conditions, compared to single phase shift modulation without the current stress optimization.