Transformer fault diagnosis based on euclidean clustering and support vector machines

A new method based on Euclidean clustering and support vector machines is presented and constructed in the paper. According to the Euclidean distances between the transformer´s state sorts, build the multi-classification model of support vector machines. The diagnosing experiments of different transformer testing scenarios show this method can avoid the blindness when building the multi-classifier, and can be used on transformer faults diagnosis commendably.