Fault Diagnosis Method of Power Transformers Using Multi-class LS-SVM and Improved PSO

We proposed a fault diagnosis method based on the multi-class least squares support vector machine(LS-SVM) and the improved particle swarm optimization(PSO) algorithm to improve the accuracy of transformer fault diagnosis. By introducing the minimum output coding, we built several two-class LS-SVMs to realize the multi-class classification of transformer diagnosis. Then we obtained the optimal parameters of LS-SVM diagnosis model using the PSO algorithm, and improved the generalization performance of this multi-class algorithm through the cross validation. Study of practical cases indicate that, after using the PSO and LS-SVM algorithm, transformer faults can be diagnosed effectively and accurately, and the accuracy is higher than that of a number of conventional transformer fault diagnosis approaches.