New Insights Into Model Predictive Control for Three-Phase Power Converters

Deadbeat control (DBC) has been investigated for decades because of its simple concept and satisfactory performance. Recently, finite control set-model predictive control (FCS-MPC) attracted wide attention in the control of power converters. Unlike DBC, FCS-MPC does not employ a modulator and directly manipulate switching states by minimizing a cost function. To improve the steady-state performance, some improved FCS-MPCs introducing two or three vectors during one control period were proposed in the existing literature. This paper proposes new insights into multiple-vector FCS-MPC for power converters by investigating its relationship with space vector modulation based DBC (SVM-DBC). Analytical studies show that two kinds of predictive methods can be unified in one framework when the control target is to minimize the tracking error. Different from enumeration-based vector selection in conventional FCS-MPC, the vector sequence and duration of the proposed multiple-vector FCS-MPC are directly reconstructed from SVM-DBC. Theoretical analysis on the state-of-the-art DBC and FCS-MPC, as well as experimental test on a three-phase two-level voltage source pulsewidth modulation rectifier confirms the effectiveness of the proposed method.