A Construction Method for the Knowledge Graph of Power Grid Supervision Business

Abstract

In the centralized supervision mode of the power grid, there are a large amount of alarm information and equipment maintenance, test and defect records. Its efficient storage and accurate retrieval play an important guiding role in the operation evaluation and supervision management of power grid equipment. This paper takes the power grid equipment maintenance, defects, alarms and oil chromatography, etc. as the research object, and proposes a top-down and bottom-up method to build a knowledge graph of power grid supervision business, which assists supervision management personnel to grasp the status of equipment in a timely manner and improve the power grid's emergency handling capabilities. Firstly, the schema layer is defined, including knowledge entities and relations between entities from top-down; then, the data layer from bottom-up is constructed through the rule-based template extraction and intelligent model-based approach; besides, the trinomial tree model is used to achieve multi-source data fusion, with a matching rate of 95.33% and an error rate of 2.3%; finally, the Neo4j graph database is applied to store and visualize the power grid supervision business knowledge graph. Also, the application of the knowledge graph is analyzed, including intelligent information retrieval and equipment condition evaluation and prediction. Through the case study of the oil chromatography data evaluation, the effectiveness of the above-mentioned knowledge graph is verified.