Current Stress Optimization-Based Unbiased Model Predictive Control for Dual Active Bridge Converters

To improve the power transfer efficiency, dynamic response, and robustness of dual active bridge (DAB) converters, this paper proposes a bias-free model predictive control (MPC) method. First, to mitigate high conduction losses and low conversion efficiency, the Karush-Kuhn-Tucker (KKT) conditions are employed to determine the optimal phase-shift angle for minimizing current stress. Then, a comprehensive set of cost functions is designed to evaluate and optimize candidate control variables, while adaptively correcting the reference commands. This approach eliminates steady-state errors caused by model parameter mismatches and enhances system robustness. Finally, an experimental prototype is developed to validate the effectiveness of the proposed control strategy.