Research on the method of constructing a knowledge map of relay protection device faults oriented to textual features.

In recent years, science and technology have been developing at a high speed, and the new generation of information technology featuring digitization, networking and intelligence has been increasingly innovative and breakthrough, pushing the whole society into the digital era. The paper, on the basis of knowledge graph technology, determines the fault classification according to the standard fault knowledge base of relay protection device, adopts the method of mapping knowledge graph ontology fields to deal with the structured data related to the faults of the relay protection device involved, extracts the text feature data information from the unstructured data related to the faults of the relay protection device through the deep learning model, focuses on the research of the vector conversion of the text feature statements, and finally extracts the information extraction model through the double neural network joint extraction of statement features. The information extraction model is constructed through the method of joint extraction of statement features by dual neural networks, and the construction of the knowledge map of relay protection device faults is finally completed through model training and result optimization. The research on the construction of relay protection device fault knowledge map for text features can improve the efficiency of abnormal data management of relay protection device, and will further promote the digital transformation of the power system.