Empowering Power Outage Prediction with Spatially Aware Hybrid Graph Neural Networks and Contrastive Learning

TMLR Paper6396 Authors

Abstract: Extreme weather events, such as severe storms, hurricanes, snowstorms, and ice storms, which are exacerbated by climate change, frequently cause widespread power outages. These outages halt industrial operations, impact communities, damage critical infrastructure, profoundly disrupt economies, and have far-reaching effects across various sectors. To mitigate these effects, the University of Connecticut and Eversource Energy Center have developed an outage prediction modeling (OPM) system to provide pre-emptive forecasts for electric distribution networks before such weather events occur. However, existing predictive models in the system do not incorporate the spatial effect of extreme weather events. To this end, we develop Spatially Aware Hybrid Graph Neural Networks (SA-HGNN) with contrastive learning to enhance the OPM predictions for extreme weather-induced power outages. Specifically, we first encode spatial relationships of both static features (e.g., land cover, infrastructure) and event-specific dynamic features (e.g., wind speed, precipitation) via Spatially Aware Hybrid Graph Neural Networks (SA-HGNN). Next, we leverage contrastive learning to handle the imbalance problem associated with different types of extreme weather events and generate location-specific embeddings by minimizing intra-event distances between similar locations while maximizing inter-event distances across all locations. Thorough empirical studies in four utility service territories, i.e., Connecticut, Western Massachusetts, Eastern Massachusetts, and New Hampshire, demonstrate that SA-HGNN can achieve state-of-the-art performance for power outage prediction.

Submission Type: Long submission (more than 12 pages of main content)

Changes Since Last Submission: In the revised manuscript, we strengthened clarity, completeness, and empirical validation in response to reviewer feedback. We clarified the distinction between static and dynamic features in the problem formulation and methodology, and refined the description of the dynamic graph learning and inference process. We added an additional case study on Eastern Massachusetts to complement the Connecticut analysis, demonstrating robustness across regions with different outage sparsity patterns. Finally, we also introduced a dedicated hyperparameter sensitivity analysis and discussed the selection of key coefficients.

Assigned Action Editor: ~Hongyang_R._Zhang1

Submission Number: 6396