A fault diagnosis model for power transformer based on statistical theory

A multi-level fault diagnosis model for power transformer fault diagnosis based on Statistical theory is presented The fault information within Dissolved Gas Analysis (DGA) is used to build fault diagnosis model and the fault diagnosis is accomplished according to the concentration distribution of typical fault gases in higher dimensional space. The proposed approach is constructing the most accuracy model from few training samples supporting, and it is very suitable to solve the problems of less typical fault data for diagnosis. The results of using the proposed model to analyze some known samples of testing data of faulty transformers show that the model possesses strong solving ability to deal with the problem. Moreover, by comparing with the traditional dissolved gas analysis methods like the neural network, there is less fault data discriminated by the proposed model and the accuracy for power transformer fault diagnosis is improved using our proposed model.