Data‐driven spatio‐temporal analysis of wildfire risk to power systems operation

Wildfires are natural or man-made disasters that continuously threaten portions of the transmission and distribution grid, thus stability electric grid. This paper presents a two-stage framework for assessing power system-wildfire risk using data-driven wildfire prediction model. The first stage estimates spatio-temporal probability potential ignition propagation deep neural network in combination with physical spread Analysis reveals similar spatial temporal patterns between model-predicted actual ignition. Motivated by these observations, second assesses grid operation terms loss load de-energisation, through combining geospatial information system data topology stochastic model developed stage. utility applications introduced proposed twofold: 1) proactive de-energisation against failure-induced wildfire, 2) spreading optimal operations exogenous wildfire. model, based on real-world dataset, is demonstrated IEEE 24-bus test mapped to study area Northern California, while results illustrate can achieve best performance detection (AUC 0.995) compared other baselines, as well demonstrates risk-aware enabled framework.