Virtual Synchronous Generator Frequency Stability Control Based on Data-Driven Model

The proposal of virtual synchronous generator (VSG) solves the problem of insufficient damping and moment of inertia in the grid-connected inverter, but when the system is disturbed, the stability of the system will be impacted. In this paper, frequency prediction base on data-driven model is added to the control module of VSG. Under the premise that the maximum capacity of energy storage release remains unchanged and the output power of the inverter does not exceed the rated capacity. Based on the limit value of the predicted frequency of the system after disturbance, and the advantage that the inverter parameters can be adjusted in real time, the setting interval of the virtual inertia and damping is increased, so as to improve the recovery effect of the output frequency of the system after disturbance. Finally, the validity of the model is verified by simulation.