Research

Publication Title: Transmission Failure Prediction Using AI and Structural Modeling Informed by Distribution Outages

Status: Complete
Publication Link: IEEE Access
Description: The Failure Prediction Model (FPM), a tree-based machine learning model, predicts transmission line vulnerabilities during extreme weather events by integrating weather, land cover, vegetation, infrastructure, and historical failure data.

Impact: This research contributed to the framework of the ISO New England (ISO) Online Weather Lookahead Study (OWLS) Model. Learn More.
The image below from ISO-NE’s Online Weather Lookahead Study tool shows, Tropical Storm Isaias barrels toward Connecticut on Aug. 4, 2020. The gray lines and red dots depict transmission lines and substations, including some connecting New England to neighboring regions. The information on the right includes weather conditions affecting a transmission line with relatively high probability of outage.

In this image from ISO-NE’s Online Weather Lookahead Study tool, Tropical Storm Isaias barrels toward Connecticut on Aug. 4, 2020. The gray lines and red dots depict transmission lines and substations, including some connecting New England to neighboring regions. The information on the right includes weather conditions affecting a transmission line with relatively high probability of outage.