引入电感参数辨识的 Vienna 整流器双矢量预测恒频控制

Finite Control set Model Predictive control (FCS-MPC) has strong robustness and fast dynamic response, so it is widely used in the control system of grid-connected inverter. However, the mismatch of inductance parameter may directly affect the observation accuracy of FCS-MPC. In addition, the switch state of traditional FCS-MPC is not fixed, which will reduce the current tracking accuracy, and even produce large current ripple and electromagnetic noise.In this paper, aiming at the problems of tracking error, difficulty in selecting the weighting factor and large searching operation for VIENNA rectifier caused by the traditional model predictive control. Based on the analysis of grid-current performance with the filter inductance parameter mismatch, a fast model predictive control strategy with constant frequency based on parameters online identification(F-MPCCF) is proposed. A single objective cost function based on the direct power is constructed by optimizing the redundant vector which is selected to balance the neutral-point potential, the design of weighting factor is avoided.The SVPWM modulation is realized by combining the zero vector adjustment method, and the smooth switching of switching state is realized. Finally, the results show that the proposed F-MPCCF has good steady-state and dynamic performance from the static, transient and neutral-point potential control.