A Novel Two-Stage Model Predictive Control for Modular Multilevel Converter With Reduced Computation

Abstract

Modular multilevel converter (MMC) has become a promising topology applied in medium/high-voltage applications due to its advantages. However, there are several challenges in the MMC, such as the appearance of harmonic circulating current. In this paper, a novel two-stage model predictive control (TSMPC) is proposed to suppress the harmonic circulating current. First, the mathematical model of the MMC is discretized to describe steady-state objectives including output currents, circulating currents, summation, and imbalance of capacitor voltages. Then, an optimal control options set consisting of inserted numbers is built by the front stage of TSMPC, while the inserted number applied to the MMC ultimately is selected from the above-mentioned set by the back stage of TSMPC. The proposed TSMPC can not only improve the control performance of harmonic circulating current but also ensure the overall control performance of the MMC system. In addition, a simplified rolling optimization method is proposed to reduce computation from the aspect of simplifying each control option prediction. An optimized search space method is also proposed to design the control options set to be evaluated from the aspect of ensuring the output current tracking and harmonic circulating current suppression. Finally, the effectiveness of the proposed method is validated by experiment results.