The Research And Application Of Transformer Data Mart

The transformer condition information should be used synthetically for conditionassessment. However, the information scattered in different departments of thecompany. Although abundant, the data is inconsistent and spread around, and theexpert knowledge available for condition assessment is very poverty. Therefore it isnecessary to build a data mart for data mining. On the view that it is hardly for users toinfer the insulation condition of the transformer out of vast data, this paper proposed adata mart framework orienting on transformer condition assessment, based on whichOLAP was done. Based on this data mart, a new method that used data miningtechnique according to DGA data was offered for transformer fault diagnosis.