Power transformer fault identification based on improved BERT-wwm-ext-CNN

Safe, economical and high quality electric transportation is the basic requirement of modern power system operation. Power transformer is an important equipment in power system. It has various types, changeable models and extensive layout. In order to achieve more comprehensive and accurate transformer fault identification and classification, this paper proposes a fault text classification algorithm based on improved BERT-CNN to. Experiments on various transformer fault data show that the model can effectively realize the accurate identification of multi-type transformer fault diagnosis.