A digital twin knowledge platform for the Yongding River Basin based on generative pre-trained large language models is designed.The platform builds a digital twin knowledge base for the Yongding River based on massive business data,books,periodicals,and other materials.Based on the knowledge base,the large language model was trained to create knowledge graphs of water conservancy objects,forecasting schemes,scheduling plans,historical scenarios,and so on.Based on the large language model,knowledge graph,and vector database,knowledge retrieval and intelligent question answering were realized,and intelligent applications for water conservancy business such as intelligent retrieval of water conservancy object information,business rule matching,historical scenario replay,and real-time scheduling plan preparation were developed.
large language modelknowledge platformknowledge graphdigital twin basin
国家科技期刊平台
NSTL
维普
万方数据
