Predictive Control of Back-to-Back Converters with DC-Link Capacitance Minimization

In this paper, we propose a predictive control algorithm for back-to-back converters that is robust to low DC-link capacitance. To achieve this goal, an additional control objective, i.e., the charging of the DC-link capacitor within each sampling period, is included in the cost function to penalize the DC-link voltage ripple. In this way, a global optimized switching state is obtained and the good performances are illustrated with small DC-link capacitance in both transients and steady state. Simulation results confirm the effectiveness of the proposed method for both, high and low DC-link capacitance.