| 作成者 |
|
|
|
| 本文言語 |
|
| 出版者 |
|
| 発行日 |
|
| 収録物名 |
|
| 巻 |
|
| 開始ページ |
|
| 終了ページ |
|
| 会議情報 |
|
| 出版タイプ |
|
| アクセス権 |
|
| Crossref DOI |
|
| 権利関係 |
|
| 権利関係 |
|
| 概要 |
Landslides present a significant geohazard risk to hydropower infrastructure. This study integrates AI-driven geospatial analysis, remote sensing, and geophysical surveys to develop an enhanced landsl...ide susceptibility mapping framework for the Manolo Fortich 1 Hydroelectric Power Plant (MF1 HEPP) in Bukidnon, Philippines. Using Slide3 AI software, InSAR, photogrammetry, and Electrical Resistivity Tomography (ERT), the study improves landslide risk predictions. First-pass AI detected 101 critical slope segments, which were refined to 83 after field validation and a second pass. Model accuracy improved, with the Area Under the Curve (AUC) rising from 0.85 to 0.91, and classification accuracy increasing from 84.42% to 91.77%. The results demonstrate that AI-based geotechnical analysis, coupled with geophysical validation, enhances prediction accuracy and reduces false positives. These findings highlight the importance of digital twin modeling in hydropower risk assessment and infrastructure resilience.続きを見る
|
Mendeley出力
2025_p0624