A Spatiotemporal Graph Neural Network Framework for Hydropower Dam State Assessment and Anomaly Identification
Wenhao Dong, Xiaoliang Feng
Shanghai Dianji University
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
This paper proposes an AST-GCN framework for hydropower dam state assessment and anomaly identification using multi-source monitoring data. The framework represents rainfall stations and the main dam as a weighted spatiotemporal graph, then combines adaptive graph convolution, bidirectional temporal-spatial attention, and multi-scale feature fusion to learn coupled operational patterns. Experiments on 365 monitoring samples from a hydropower station in 2025 show 98.7% stateassessment accuracy, 97.2% anomaly-detection F1 score, 99.1% AUC-ROC, and inference latency below 65 ms. The results indicate that AST-GCN can support real-time intelligent dam operation and maintenance.
逐年被引趋势
暂无年度引用数据
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
工程Dam Engineering and Safety
Advanced Graph Neural Networks · Environmental Monitoring and Data Management