Efficient Model Predictive Control with Fixed Switching Frequency Computation of Virtual Synchronous Generators

This essay presents an improved model predictive control (MPC) method with fixed switching frequency (FSF) for midpoint-clamped three-level virtual synchronous generator control algorithms. Compared with the traditional MPC used for virtual synchronous generators, The computational burden is notably minimized by the proposed MPC while balancing the midpoint potentials of the dc-side capacitors, and reduces the common-mode voltage (CMV) amplitude by dynamically selecting the optimal voltage vector. In the improved MPC, three voltage vectors are used per control cycle for seven-segment allocation modulation to achieve a fixed switching frequency. The viability of the proposed predictive control approach is validated by assessing the system’s steady-state and dynamic performance via SIMULINK-based simulations.