Virtual-multilevel Model Predictive Control with Error Feedback for Modular Multilevel Converters

Huancheng Lin,Zhixin Wang
DOI: https://doi.org/10.1049/iet-gtd.2016.1418
2017-01-01
IET Generation Transmission & Distribution
Abstract:This study proposes a stationary coordinate system-based virtual-multilevel model predictive control strategy with error feedback for the control of the modular multilevel converter (MMC). In order to reduce the number of considered states for the ac-side current control, a cost function is established according to the mathematical model of the MMC in the two-phase stationary coordinate system. The selection of the output voltage levels is optimised with the introduction of the virtual-multilevel algorithm. By using real-time submodule (SM) capacitor voltages in current prediction, the circulating current control is combined with the SM capacitor voltage sorting to improve the control accuracy of the circulating current. This study also proposes an error feedback correction method to eliminate the steady-state error in current control and avoid parameter sensitivity of the control performance. The simulation results and experimental results from a down-scaled MMC prototype are provided to verify the validity and efficiency of the proposed method.
What problem does this paper attempt to address?