In order to improve the efficiency of transmission line fault inspection,operation and maintenance,and the level of line intelligence,a knowledge graph-based transmission line fault identification and location positioning technology was proposed.Firstly,entity extraction was carried out using long and short memory networks and conditional random fields.Relationship extraction used dependency syntax analysis to obtain corresponding triplet structured data,and event extraction was carried out using an event framework based method;secondly,the ontology layer and data layer were constructed using a combination of text and instance methods,as well as an entity alignment probability model based method,to achieve rapid query of transmission line fault causes and accurate positioning of fault locations;finally,an intelligent question answering system was built on this basis.The proposed technical system can quickly analyze equipment faults related to transmission lines,improve the efficiency of transmission line fault inspection and operation maintenance,and ensure the reliability of transmission line operation.
关键词
输电线路/知识图谱/长短记忆网络+条件随机场/故障识别与故障定位/智能问答
Key words
transmission line/knowledge graph/long short-term memory+conditional random field/fault identification and location/intelligent question answering
维普
万方数据