Performance Improvement of Model Predictive Control for Modular Multilevel Converters by Auto-regulating the Weighting Factor Value

The Model Predictive Control (MPC) has attracted much attention for controlling power electronic topologies, such as Modular Multilevel Converters (MMC). Despite numerous advantages of MMC, several control objectives, i.e. circulating and output currents control, and also the cell capacitors' voltage balancing should be realized. Although implementing several control objectives with a single cost function is an inherent feature of the MPC scheme, choosing appropriate weighting factor values is an important issue, besides the well-known high computational load problem. In this paper, a technique is proposed for adjusting the weighting factor value, in a way that the output current THD and also the circulating current RMS values are simultaneously minimized. This is done by designing a simple look-up table and also using sgn functions, in order to auto-update the weighting factor value during the converter operation. Finally, the performance of the proposed method is validated through conducting some simulations in the MATLAB/Simulink environment.