Model Predictive Control Strategy of Multi-Vector Fixed Frequency for Vienna Rectifier

Model predictive control (MPC) has been recognized for its potential in enhancing the control performance of Vienna rectifiers. However, the conventional finite-set MPC is associated with significant drawbacks, including high grid-side current harmonics, an unfixed switching frequency, and a substantial computational burden for vector selection, which can adversely affect the control performance and efficiency of Vienna rectifiers. To address these issues, a three-vector fixed-frequency MPC algorithm based on sector identification has been proposed. In this approach, deadbeat control is employed to derive voltage control from the reference current, followed by the identification of sector-active vectors. A cost function is constructed using the predicted current and midpoint potential to determine the vector operation time. Simulation and experimental studies, including tests on a 10 kW prototype, have been conducted. The results demonstrate that the proposed method can effectively fix the switching frequency, mitigate midpoint-voltage oscillation, and enhance the dynamic response of the controller.