Chinese named entity recognition of transformer bushing fault based on BiLSTM-CRF

Bushing is an important part of transformer, which is often in the environment of high voltage and high current, and is prone to failure. The vast majority of transformer failures are caused by bushing failures, which will bring about serious economic losses. With the development of information technology, the power system is also moving towards digitalization. Knowledge graph plays an important role in the rapid processing of power information. For this technology, named entity recognition is the key step to build knowledge graph, which can extract power information entities and promote the process of power system digitalization. Therefore, this paper proposes a transformer bushing fault extraction method based on Chinese named entity recognition, and extracts the fault information from the bushing fault text based on the BiLSTM-CRF natural language processing model. First of all, according to the characteristics of transformer bushing fault text, the schema layer is constructed using the top-down entity extraction method, and its data layer is also constructed according to the attributes defined in the schema layer; Then, the mainstream machine learning models are compared with the model proposed in this paper, and the effectiveness of this method is evaluated by using precision, recall, F1 score, which reaches 92.91%, 91.89% and 92.40% respectively.