Selection of Proper Non-linear Kernel Parameter in Support Vector Machine Algorithm for Classifying the Internal Fault Type in Winding Power Transformer

Abstract

This paper proposes the proper kernel function based on a support vector machine (SVM), which is used to classified the internal fault type in power transformer. The gaussian kernel and polynomial kernel that is two type of non-linear kernel parameter, is compared in terms of the average accuracy and time of training processing. The results shown that the polynomial kernel parameter of SVM algorithm able to classified the internal fault type with satisfactory accuracy while it use the average time less than the other. Benefit of polynomial kernel can be applied for fault diagnosis in future.