Model Predictive Control with Active Sensor Noise Suppression for Dual Active Bridge Converter

Dual Active Bridge (DAB) converters are widely used in energy storage systems, fuel cell systems and microgrids for its excellent characteristics. Recently, model predictive control (MPC), which is known for its high dynamic performance and multi-objective optimization capability, has been applied to DAB converters. MPC relies on feedback variables to achieve state variable prediction, where the sensor sampling noise will degrade MPC performance and greatly limits its application. In this work, we analyze the effect of sensor noise on the MPC performance of a DAB converter, and propose an MPC method with active sensor noise suppression. The proposed method independently sets the MPC discretization cycle and the switching cycle, so that the control performance deterioration can be suppressed without reducing the switching frequency or adding filters. The simulation results verify that compared with the classical MPC method, the proposed method can effectively reduce the inductor current stress and the output voltage ripple, suppress the transient bias, and improve the efficiency of the DAB converter, when the sensors suffer from sampling noise.