Model Predictive Control of MMC-Based Medium Voltage Microgrid for Grid Connected and Islanded Operation

This article focuses on the development of a model predictive control (MPC) for a modular multilevel converter (MMC)-based medium voltage microgrid. The developed MPC is aimed to effectively control MMC-based microgrid during both grid-connected (GC) and islanded (IS) modes, with seamless transitions between the two. The developed MPC anticipates state variables behavior and uses a cost function to determine an optimal switching sequence. As a result, unlike classical vector control methods, the developed MPC eliminates need for pulse width modulation schemes, proportional-integral controllers, and extensive tuning efforts. The developed MPC enhances system performance by incorporating power quality improvement and AC fault ride-through capabilities. Additionally, the developed MPC also provides a submodule fault-tolerant feature to ensure system reliability and stability. The effectiveness and robustness of the developed MPC are validated through MATLAB/Simulink-based simulations. Further, extensive experimental studies are performed on a scaled-down laboratory prototype of a 5-level MMC, using dSPACE 1202. The performance evaluation encompasses various operating conditions, including GC and IS operations, smooth transitions between these modes, AC fault ride-through, unbalanced and distorted grid conditions, submodule fault and power quality enhancement. The performance of the developed MPC is compared with traditional vector control methods, demonstrating its superiority in managing medium-voltage microgrid operations.