Combining Remotely Sensed Imagery with Survival Models for Outage Risk Estimation of the Power Grid
Abstract
Vegetation management of power grids is essential for reliable distribution of services, prevention of forest fires and disruption of electricity due to tree fall. In this paper, we introduce a vegetation analysis system that utilizes information from GIS data, aerial and satellite imagery to estimate vegetation profile within a buffer zone. This vegetation profile is further combined with operational parameters of the grid to develop a survival model which predicts the outage risk of a power-line in an electrical grid. Using historical data, we show that the risk scores thus obtained are significantly better at developing trimming schedules for grid power-lines, compared to existing available methods.
Cite
Text
Jain et al. "Combining Remotely Sensed Imagery with Survival Models for Outage Risk Estimation of the Power Grid." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021. doi:10.1109/CVPRW53098.2021.00131Markdown
[Jain et al. "Combining Remotely Sensed Imagery with Survival Models for Outage Risk Estimation of the Power Grid." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021.](https://mlanthology.org/cvprw/2021/jain2021cvprw-combining/) doi:10.1109/CVPRW53098.2021.00131BibTeX
@inproceedings{jain2021cvprw-combining,
title = {{Combining Remotely Sensed Imagery with Survival Models for Outage Risk Estimation of the Power Grid}},
author = {Jain, Arpit and Shah, Tapan and Yousefhussien, Mohammed and Pandey, Achalesh},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2021},
pages = {1202-1211},
doi = {10.1109/CVPRW53098.2021.00131},
url = {https://mlanthology.org/cvprw/2021/jain2021cvprw-combining/}
}