Research on Text Classification Method of Distribution Network Equipment Fault based on Deep Learning

摘要

With the development of science and technology, the intelligent distribution network is making rapid progress and the demand for power supply is also increasing. The classification and analysis of the power outage causes in distribution network is helpful to the prediction of power failure. There is a large gap in the proportion of categories of the sample data of power grid outage causes, and serious category imbalance will lead to the deviation of classification results. In order to classify the original data more reasonably, the sample value added technology based on generative adversary network is used to expand the category data with less data in this paper. In order to realize automatic classification of power grid outage causes, the method of text data preprocessing and the text classification model based on deep learning are studied. Finally, the experimental results show the influence of sample capacity and classification accuracy.