Power Transformer Fault Diagnosis Based on Ensemble Learning

In the aspect of transformer fault diagnosis, the relationship between transformer fault and dissolved gas in oil has been particularly described in this paper. Considering the objective fact that transformer fault data is far less than normal data, the balanced processing method of unbalanced data sets in the classification process has been discussed. Considering these factors, all kinds of fault state data similar to the normal state data were selected as sample data, and ensemble learning was used to fault diagnose the transformer. The experimental results show that the method used in this research has an accuracy of 94.5% in fault diagnosis, which is significantly higher than other fault diagnosis methods, verifying the correctness and feasibility of this method.