Transformer Fault Diagnosis Based on BOA Optimized SVM

To improve the classification accuracy of transformers in different operating states, a BOA-SVM classification method for transformers in different operating states is proposed. Due to the excellent global search capability of the BOA method, in order to address the issue of the significant impact of key parameters on transformer operation status classification results in the SVM method, BOA is used to determine the key parameters of the SVM method, and the BOA-SVM model is applied to transformer operation status classification. In the end, the BOA-SVM fault diagnosis model has higher classification accuracy compared to using only the SVM model, with an average classification accuracy of 95.9%, and can effectively diagnose and classify the operating status of transformers.