Framework and Key Technologies of Knowledge-graph-based Fault Handling System in Power Grid

In order to deal with the increasingly complicated situation of safe operation in power grid, it is one of the key technologies to model, manage and apply the knowledge hierarchy to improve the intelligent level of power dispatching. This paper proposed a framework of knowledge- graph-based fault handling system, in which the unstructured data of the operation rules, emergency plans and detailed rules with a large number of texts were converted to structured knowledge graph that can be represented, operated and reasoned. The construction and application of the presented knowledge graph contain five levels, i. e., basic data, graph construction, information analysis, reasoning decision and human-computer interaction layer. According to the scenarios and requirements of power grid fault handling, the domain knowledge graph is divided into four parts, i. e., graphs for power equipment, concepts, operation rules and fault cases. The detailed solutions and key technologies were analyzed to realize knowledge-graph-based fault information analysis, intelligent decision- making and human-computer interaction. Finally, the future application and challenge of the knowledge graph in intelligent assistant decision for power scheduling were summarized and forecasted. © 2020 Chin. Soc. for Elec. Eng.