Research and application of power transformer technical standard retrieval algorithm based on knowledge graph

Abstract

In view of the difficulty in using unstructured text data and the shallow application of equipment knowledge in the process of transformer equipment operation inspection by power supply companies, this paper proposes a knowledge map technology framework to support transformer intelligent management by using knowledge map, natural language processing and other artificial intelligence technologies. At the same time, this paper proposes a method to screen new knowledge through text error checking and a method to screen degraded knowledge through text path searching. After the knowledge map is updated, the integrity of knowledge is verified through text error checking. Finally, this method is verified in the scenarios of flexible question answering of power transformer equipment information and automatic extraction of transformer fault reports. The experimental results show that this scheme improves the accuracy of the knowledge map, can properly integrate new knowledge into the knowledge map, and effectively improves the application effect of the defect text knowledge map.