Research on intelligent decision-making system for coal mine water hazard early warning and emergency response based on knowledge graph

WENG Mingyue; XU Hua; YIN Huimin; FU Ensan; GAO Wenzhong; YE Lan; SHAN Yao

doi:10.11799/ce202602013

Abstract：Water hazard accidents are among the major disaster types in coal mine production.Although certain knowledge and technologies have been accumulated for coal mine water hazard prevention,deficiencies remain in deep data mining,systematic application of empirical knowledge,and intelligent decision support.To ensure safe production in coal mines and enhance the informatization and intelligence levels of water hazard prevention,control,and rescue,we systematically review the technical architecture of intelligent early warning systems and decision support systems for coal mine water hazards based on knowledge graphs.The construction of a coal mine water hazard knowledge graph primarily involves two core components:ontology and relationship extraction based on the BiLSTM-CRF model,and the establishment of a graph database.Building upon the graph database,an intelligent question-answering system integrating naive Bayes-based question classification,question parsing,and graph database querying is designed,enabling real-time and efficient retrieval of coal mine water hazard accident cases.Furthermore,a framework for an intelligent early warning platform for coal mine water hazards is constructed,encompassing the physical layer,transmission layer,data processing layer,and application layer.The integration methods of the knowledge graph with text libraries,model libraries,and on-site monitoring data are analyzed in depth.A comprehensive decision support system is formed,covering data acquisition and analysis,decision support solution generation,and real-time generation of emergency response plans.The research findings provide a theoretical reference for further enhancing the informatization and intelligence levels of coal mine water hazard prevention,control,and emergency rescue.