Research on Model Predictive Control Method for Dual Active Bridge Based on Data-driven Predictive Model

This paper introduces a model predictive control (MPC) method for dual active bridge (DAB) based on a data-driven model. By collecting historical operational data from the DAB, a data-driven model for the DAB can be obtained through training. This model exhibits high-precision predictive capabilities for a single time step. We utilize the predicted data for rolling optimization calculations to solve the cost function, aiming to achieve optimized control effects