A generative model for synchrophasor network traffic in smart grids

Smart grid technologies can revolutionize power distribution. However, engineers can’t write the software to realize this goal without large, publicly available datasets. Power companies are reluctant to share the data they collect because of consumer privacy concerns and concerns that they will expose themselves as targets of cyber attacks. We introduce a generative model for one of the most common types of smart grid data: synchrophasor network traffic. Our model can generate an unlimited amount of smart grid data while mitigating risks associated with exposing consumer and network information. Engineers can use our model to prototype smart grid software such as anomaly detection and state estimation.