Virtual Vector Modulation-Based Model Predictive Control Strategy for Virtual Synchronous Generators

To address the challenges of insufficient system inertia and harmonic distortion in renewable-dominated power systems, this paper proposes a dual-loop Virtual Vector Modulationbased Model Predictive Control (VMPC) strategy for photovoltaic (PV)-storage grid-connected systems. The outer power loop employs Model Predictive Control (MPC) to regulate Virtual Synchronous Generator (VSG) reference power according to synchronous generator rotor dynamics. The inner current loop adopts an Discrete Vector Set Model Predictive Current Control (DVSMPCC) with $\mathbf{1 3}$-vector Discrete Space Vector Modulation (DSVM), achieving precise tracking of the outer loop reference voltage. Simulation and hardware-in-the-loop experiments verify that the proposed strategy: 1) Improves frequency regulation accuracy with $\lt 0.1 \mathrm{~Hz}$ deviation during $40 \%$ load steps; 2) Reduces grid current THD to $\mathbf{1. 7 \%}$ (vs. $\mathbf{3. 8 \%}$ for conventional FCS-MPC); $\mathbf{3}$) Maintains stable operation under 0.05 pu voltage sags. The DSVM-based EVS-MPCC demonstrates superior performance in mitigating current ripples and fixed-order harmonics compared to traditional methods.