Diagnosis and Prediction of Transformer Faults of Support Vector Machine Optimized by Particle Swarm Algorithm

Abstract

Transformer as the hub of electrical power system plays a notable role in whole power grid; hence the failure of transformer may cause huge loss and threaten the life and property security of people. The diagnosis and prediction of transformer faults is quite necessary and important. To perfect the diagnosis and prediction of transformer faults, this study developed an optimized support vector machine (SVM) based on particle swarm optimization (PSO) fault diagnosis and established diagnosis and prediction models and performed optimization selection on the parameters of the models. Finally it was verified that PSO based SVM had a high preciseness in diagnosing and predicting transformer faults, suggesting a favourable practicability.