Model Predictive Control for Midpoint Potential Balance of T-type Three-level Bidirectional Converter Based on Redundant Vector

The T-type three-level bidirectional converter is suitable for application in low-voltage AC/DC hybrid distribution network interconnection scenarios due to its advantages of low loss and high output power quality. In view of the imbalance of midpoint potential on the DC side of T-type three-level bidirectional converter, model-predictive controls have good application value because it is easy to realize multi-objective optimization. However, the traditional finite set model predictive control (FCS-MPC) has the problem of difficulty in weight factors adjusting, which affects the midpoint potential control effect. To solve the problems above, a model-predictive control for midpoint potential balance based on redundant vectors is proposed, which uses the characteristics of redundant small vectors to realize the control of midpoint potential balance and avoid the selection of weight factors. On this basis, the cost function is used to calculate the action time of the vector and predict the optimal switching frequency to improve the steady-state performance of the system. Finally, the effectiveness of the proposed strategy is verified through experimental testing.