Abstract:The building of a knowledge graph within the water conservancy sector can offer structured knowledge support for digital twin water conservancy projects. Based on the forecast and scheduling, engineering safety and business rule data of Xiaolangdi and Wanjiazhai key water control projects on the Yellow River, an entity recognition and relationship extraction model was used to build a water conservancy knowledge graph for digital twin engineering (KG4DT), and the graph was visualized and displayed. It determined the entity classification system and relationship classification system with the key water control project as the core, and built the pattern layer of KG4DT. It used a secondary development annotation platform to perform semiautomatic annotation and manual proofreading on unstructured text and built the data layer of KG4DT. Taking the flood control, sedimentation reduction, water supply and power generation businesses of the Xiaolangdi and Wanjiazhai key water control projects on the Yellow River as the main line, it combined the technologies such as AC automaton to design knowledge retrieval and automatic generation of contingency plans, which could provide intelligent decisionmaking support for engineering operations.
Abstract: The building of a knowledge graph within the water conservancy sector can offer structured knowledge support for digital twin water conservancy projects. Based on the forecast and scheduling, engineering safety and business rule data of Xiaolangdi and Wanjiazhai key water control projects on the Yellow River, an entity recognition and relationship extraction model was used to build a water conservancy knowledge graph for digital twin engineering (KG4DT), and the graph was visualized and displayed. It determined the entity classification system and relationship classification system with the key water control project as the core, and built the pattern layer of KG4DT. It used a secondary development annotation platform to perform semiautomatic annotation and manual proofreading on unstructured text and built the data layer of KG4DT. Taking the flood control, sedimentation reduction, water supply and power generation businesses of the Xiaolangdi and Wanjiazhai key water control projects on the Yellow River as the main line, it combined the technologies such as AC automaton to design knowledge retrieval and automatic generation of contingency plans, which could provide intelligent decisionmaking support for engineering operations.
Key words: water conservancy knowledge graph; digital twin; knowledge description system; application