Construction and application of power transformer fault knowledge graph based on Sentence-BERT

With the rapid development of smart grid, the relevance of multidimensional information in power grid is weak, and the decision-making generation efficiency in operation and maintenance process is low. At present, knowledge graph has innovative applications in various fields, which improves the efficiency of knowledge query in related professional fields. This paper proposes a power transformer fault diagnosis system driven by knowledge graph, which aims to solve the problems of weak information relevance and low decision-making efficiency in transformer operation and maintenance; The system is composed of three modules: data acquisition, data analysis and data service. By constructing the knowledge graph and model of fault diagnosis, the intelligent query and diagnosis of transformer fault are realized; The construction of knowledge graph is based on expert knowledge and fault data, and the entity extraction is carried out by using the Sentence-BERT-BiLSTM-CRF model, which improves the extraction accuracy of transformer fault information; The three tuple data is stored in the neo4j graph database, and the cypher language is used for efficient query, providing accurate fault diagnosis and maintenance strategies for operation and maintenance personnel. Experimental results show that the proposed method can effectively improve the accuracy and efficiency of fault diagnosis.