Multistage Model Predictive Control for Modular Multilevel Converter

With the attractive advantages of modularity, low harmonic and redundancy control, the modular multilevel converters (MMC) has received much attention and research in the field of medium and high voltage. In this paper, a multistage model predictive control method (MMPC) is proposed for MMC. The proposed MMPC method contains three stages: 1) the front stage of MMPC obtains the reference control options for output current tracking; 2) the intermediate stage of MMPC designs the control options set according to circulating current suppression; 3) the back stage of MMPC determines the final applied control options. Finally, the optimal switching states are assigned to the specific cell based on sorting algorithm. Compared to the conventional hierarchical MPC method, the proposed MMPC method provides an improved performance on circulating current suppression on guarantee of the output current tracking. Finally, the simulation results verify the effectiveness of the proposed MMPC method.