A Knowledge Graph Construction Method for Substation Infrared Operation and Maintenance Data Retrieval and Statistical Analysis

The infrared operation and maintenance report of electrical equipment contains a large amount of equipment fault data, such as text, images and so on. However, these heterogeneous data are difficult to be analyzed and utilized efficiently by operation and maintenance personnel due to the lack of structured storage management. In this paper, the method of deep learning combined with knowledge graph technology is used to realize the knowledge extraction, storage and application of infrared operation and maintenance data of electrical equipment, and improve the efficiency of the use of heterogeneous data in the report. Firstly, the infrared operation and maintenance report information is analyzed to construct the overall framework of the top-down knowledge graph. Then, the improved BERT-BiLSTM-CRF entity identification method is used to entity extraction of the report. Finally, the extracted entities and relationships are stored in the Neo4j graph database, and the infrared operation and maintenance knowledge graph visualization interface is designed through the Neovis.js visualization library, which realizes the interactive functions such as fault equipment information query, original report viewing and equipment fault statistics. And supported the efficient retrieval and statistics of infrared fault information by electrical equipment operation and maintenance personnel.