Study on Complex Kernel Function and Parameter Optimization of SVM in Transformer Fault Diagnosis

Abstract

In order to make transformer potential fault diagnose effectively, Support Vector Machine (SVM) is introduced as an effective algorithm. Firstly, the above SVM algorithm is formed by four common kernel functions: linear kernel function, polynomial kernel function, RBF kernel function and Sigmoid kernel function, Secondly, Differential Evolution Algorithm (DEA) based on new fitness function is introduced to solve the problem of parameter selection. The simulation results show that, with choosing best parameters under DEA, the complex kernel function SVM makes the fault diagnosis accuracy increase up to 94.44%.