Fault diagnosis of transformer based on cluster analysis

In order to solve the problem of the imbalance between the fault data and the normal ones in the fault diagnosis of transformer, we adopt k-means algorithm to cluster the data. The result of clustering shows the existence of the boundary class that is between fault and normal ones. The separation of boundary class from the fault data and the normal ones improves the reliability and early warning ability of fault diagnosis of transformer, as well as reduces the influence from the imbalance of the two kinds of data.