.png){kind=icon}
.png){kind=icon}
Return
Integrating knowledge graphs and LLMs for intelligent full-chain safety risk pre-control in large hydropower projects
.png){kind=icon}
OA
AI
Y
K
S
P
Z
DOI：10.1016/j.autcon.2026.107194.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Developed a full-chain “risk–hazard–accident” framework for CSR pre-control in LHPs. • Proposed an ALBERT-BiLSTM-Att-CRF model for domain-specific CSR entity recognition. • Constructed a CSR knowledge graph with 1853 entities and 4726 relationships. • Integrated the KG with an LLM for dual-mode risk pre-control decision support. • Achieved 0.8667 closed-ended QA accuracy and 4.21/5.00 expert relevance score.
Keywords:
Large hydropower projects
Construction safety risk
Knowledge graph
Large language model
Risk pre-control
Decision support system
Journal
IF:
11.5
Papers:
6.1K
Citations:
4.2W
Organization
Cited Papers