Predicting Power Outage During Extreme Weather Events with EAGLE-I and NWS Datasets

Abstract

Extreme weather events, such as hurricanes, severe thunderstorms, and floods can significantly disrupt power grid systems, leading to electrical outages that result in inconvenience, economic losses, and life-threatening situations. There is a growing need for a robust and precise predictive model to forecast power outages, which will help prioritize emergency response before, during, and after extreme weather events. In this paper, we introduce machine-learning models that predict power outage risk at the state level during and after extreme weather events. We jointly utilized two publicly available datasets: the U.S. historical power outage data collected by the Environment for Analysis of Geo-Located Energy Information (EAGLE-$\mathrm{I}^{\mathrm{T}\mathrm{M}}$) system, and the National Weather Service historical weather alert data sets. We highlight our initial result and discuss future work aimed at enhancing the model’s robustness and accuracy for real-world applications.