Transformer fault diagnosis based on support vector machine

Analysis of dissolved gases content in power transformer oil is very important to monitor transformer latent fault and ensure normal operation of entire power system. Analysis of dissolved gases content in power transformer oil is a complicated problem due to its nonlinearity and the small quantity of training data. Support vector machine (SVM) has been successfully employed to solve classification problem of nonlinearity and small sample. However, SVM has rarely been applied to diagnosis transformer fault by analysis the dissolved gases content in power transformer. In this study, support vector machine is proposed to analysis dissolved gases content in power transformer oil, among which cross-validation is used to determine free parameters of support vector machine. The experimental data from the electric power company in Sichuan are used to illustrate the performance of proposed SVM model. The experimental results indicate that the proposed SVM model can achieve very good diagnosis accuracy under the circumstances of small sample. Consequently, the SVM model is a proper alternative for diagnosing power transformer fault.