An Improved Model-Free Predictive Current Control Method for Vienna Rectifier in Charging Pile for Evs

In electric vehicle (EV) DC charging pile systems, most Vienna rectifiers are used as the front-end structure. The three-phase three-level Vienna rectifier offers benefits such as reduced switching stress and improved power factor, and is used in high-power DC charging piles. Traditional model predictive current control (MPCC) is very sensitive to changes in machine parameters. When the model parameters and the real machine parameters do not match, the system performance will be significantly reduced. This paper proposes an improved model-free predictive current control (MPFCC) to solve this problem. It uses the current information of the past two moments to realize online updating of the model gain and dynamic part, and further obtains the switch driving signal through space vector modulation (SVM) and an expanded state observer (ESO) is introduced to suppress disturbances. This control technique enhances the resilience of the control system, reduces the calculation time, and delivers superior results regarding both current harmonics and dynamic overshoot. The proposed method was compared with traditional MPCC in MATLAB/Simulink, and simulation and experimental results verified the effectiveness of the method.