An Optimal Double-Vector Finite Set Model Predictive Control Strategy Based on Deadbeat for Vienna Rectifier

The traditional finite set model predictive control for Vienna rectifier takes the minimum cost function as the control goal and outputs a single optimal vector by traversing the set of candidate vectors. However, it can cause great power fluctuation and high input current distortion. To solve these problems, this paper proposes an optimized double-vector model predictive control strategy for Vienna rectifiers. This strategy outputs two action vectors in the control cycle to jointly track the reference, and the action time of the two action vectors is calculated by the idea of deadbeat. In order to reduce power fluctuations during the control process, the sequence of vector actions within adjacent control cycles and control periods is optimally adjusted. At the same time, the neutral point (NP) potential sign on the DC side is pre-judged to screen the redundant small vectors involved in the cyclic calculation to balance NP voltage on the DC side. Simulation and experimental results prove that the proposed control strategy can effectively reduce the power fluctuation, relieve control system dependence on sampling frequency, improve the quality of the input current and balance the NP voltage in the DC side.