Construction and Application of a Knowledge Graph in the Field of Fully Automated Metro Train Degraded Operation

In the context of the degraded operation of fully automated trains, the identification and management of risk factors become challenging due to the vast and intricate textual data associated with faults that lead to various contingencies. Consequently, to enhance the management of fault data, this paper employs the methods of information extraction, knowledge processing, and reasoning to establish a structured/unstructured text-oriented system to analyze the model. In this study, we focus on a specific rail line as the subject of investigation. The extraction of core knowledge entities and relationships in fault logs and fault texts is accomplished through the BERT model, while a knowledge graph of degraded train operation in urban rail transit is constructed using tools such as Neo4j, which effectively visualize and display the semantic associations between emergency events and fault information in subway train operation.