A Survey on Application of Data Mining on Transformer Condition Assessment

People’s attention of transformer condition assessment is increasing in recent year and data mining is applied in transformer condition assessment with its obvious superiority in dealing with complex data and finding potential problems. In this paper, the process of the research on applying data mining to transformer condition assessment is summarized. The research results of data processing methods such as rough set theory, principal component analysis and colony algorithm as well as the research result of transformer condition assessment such as vector machine, neural network, Bayes network, association rules and fuzzy theory are introduced, analyzed and compared in detail. The advise on the choice and quantification of figure, the employ of the pattern recognition method, the blend of multi-information and visualization display are given.