Transformer Fault Diagnosis Based on Fuzzy Support Vector Machines

Abstract

Due to lack of typical damage samples in the transformer fault diagnosis, a new fault diagnosis method based on fuzzy support vector machines (FFSVMs) was presented. According to the method, the five characteristic gases dissolved in transformer oil were extracted by the K-means clustering (KMC) method as feature vectors, which were input to fuzzy optimal multi-classified SVMs for training. Then the FSVMs diagnosis model was established to implement fault samples classification. Experiment showed that by adopting facture extraction with KMC, the diagnosis information was concentrated and the consuming in parameter determination was solved effectively. The presented method enabled to detect transformer faults with a high correct judgment rate, and can be used as an automation approach for diagnosis under condition of small samples.