Fault diagnosis and system development of power transformer based on support vector machine

Abstract

Fault diagnosis of power transformer is important for safety of the device and relevant power system. In the study, support vector machine(SVM) classifiers combined with the form of binary tree are applied to construct diagnostic model of power transformer, and the diagnostic system structure of power transformer is presented on the basis of the model. SVM is a novel machine learning method based on SLT. It is powerful for the practical problem with small sampling, nonlinear and high dimension, which is very suitable for online fault diagnosis of transformer. The test results show that SVM has higher diagnostic accuracy than BP, IEC three ratios in fault diagnosis of power transformer.