Fault model construction based on gas chromatography of insulation oil by data mining technique

By DataCruncher, the gas chromatography data of a series of transformers are collected and processed. The data are mined. The fault model of the transformers is constructed. The relationship between the contents of the dissolved key gases and other summed parameters and the occurrence of fault in transformers is formulated. It approved that by this model it can give high precision of prediction in transformer fault diagnosis. The predicted result is agreement other methods.