Support Vector Machine Transformer Fault Diagnosis Based on Rough Sets and Cuckoo Search Algorithm

Abstract

Since the traditional transformer fault diagnosis methods do not reflect the working status of the transformer comprehensively and the error is relatively large, for this reason, a diagnosis method that combines rough set with support vector machine and adds cuckoo search algorithm to it is proposed. Firstly, the rough set is used to carry out data approximation of the original data, remove the redundant data, extract the feature information, and then use the cuckoo search algorithm to get the best parameters of the support vector machine to classify the feature information and get the type of transformer fault. It is proved that this method has the advantages of convenient operation, fast diagnosis speed and high classification accuracy than the traditional method.