A Triple-Voltage-Vector Model-Free Predictive Current Control Scheme for Voltage Source Inverters

Abstract

Model-free predictive current control (MFPCC) method is a promising method to enhance the robustness of model predictive current control. However, conventional MFPCC methods apply only a voltage vector per control period, resulting in a large output ripple. They also have updating stagnation issue which affects the accuracy of current gradient and degrades the performance of voltage source inverters (VSIs). This paper proposes a triple-voltage-vector (TVV) MFPCC strategy for VSIs based on an advanced current gradient updating method. In the proposed strategy, a TVV-MFPCC is developed to improve current performance and enhance parameter robustness. Furthermore, an advanced current gradient updating method is proposed to eliminate stagnation effect of current gradient by considering the relationship between current gradient of different VVs. Simulation and experimental comparisons on a VSI assess the performance of the proposed TVV-MFPCC scheme.