Transformer Fault Diagnosis Based on Seagull Algorithm Optimized Support Vector Machine

Abstract

Aiming at the problem of low accuracy of support vector machine for transformer fault diagnosis, a seagull optimization algorithm support vector machine method is proposed. Since the original features of the faulty transformer are less, firstly add different gas fraction ratio features, increase the information features contained in the transformer fault data, and then use principal component analysis (PCA) to extract the input variable features and reduce the dimension of the feature variables. Reduce the correlation between variables, and finally use the seagull optimization algorithm (SOA) to optimize the parameters of the support vector machine. The simulation results show that compared with particle swarm optimization (PSO) and genetic algorithm (GA), the seagull optimization algorithm optimized support vector machine (SOA-SVM) can significantly improve the accuracy of transformer fault diagnosis, and the reliability and generalization performance are also improved.