Fault diagnosis model of power transformer based on grey relation with combined weights

Dissolved gas analysis (DGA) is an important method to identify the transformer faults. In this paper, the grey relation theory based on the DGA data is introduced, and a grey relation analysis method based on the combined weights for diagnosing transform faults is proposed. For judging a transformer fault type, different dissolved gases are given different weights. Experts determine the subjective weight of each dissolved gas with AHP in diagnosing transformer fault type, however there are too many subjective factors with this method. Therefore, the grey relation analysis is introduced in this paper in order to avoid non-uniform weight distribution caused by experts' subjective factors, and objective weight of each indicator is determined by using of entropy method. Thus, the combined weights of indicators are obtained and applied to the proposed method for transformer fault diagnosis. Actual case study is performed with this method, and diagnosis results verify its effectiveness.