深層学習による台風時の電柱二次被害予測の検討

カテゴリ: 部門大会

論文No: OS3-7

グループ名: 【C】2024年電気学会電子・情報・システム部門大会

発行日: 2024/08/28

タイトル(英語): A Study of Indirect Damage Prediction for Electrical Poles During Typhoons Using Deep Learning

著者名: 横山 和弘（関西電力送配電株式会社）,森 大樹（SAS Institute Japan株式会社）

著者名(英語): Kazuhiro Yokoyama (Kansai Transmission and Distribution,Inc.),Hiroki Mori (SAS Institute Japan)

キーワード: 台風|深層学習|被害予測|二次被害|配電設備電柱|Typhoon|Deep learning|Damage prediction|Indirect damage|Distribution equipmentElectrical Poles

要約(日本語): TTyphoon No. 21 hit Japan in 2018, and more than 1,000 electrical poles in Kansai's distribution lines were damaged by strong winds, causing large outages. Most of the damage to electrical poles was caused by flying objects blown down by strong winds, which contacted with electrical poles and wires.
Prediction of damage to distribution facilities caused by typhoons is extremely important to consider damage mitigation measures, such as reviewing design standards, making restoration plans, and preparing restoration systems and materials.
In this study, a prediction model of damage to electrical poles caused by flying objects was considered to improve the accuracy of predicting damage to electrical poles caused by typhoons.