Transformer defect correlation analysis based on Apriori algorithm

The association rule from data mining technology was applied into transformer defect analysis so that the frequent pattern, the dependency and the causality between classification and decision attributes could be found based on data of defects. As a result, correlation properties among grid fault elements were seized macroscopically. In this paper which focused on the frequent item mining algorithm research for transformer defect correlation analysis, the definitions related to the association rule were introduced. Specific to weaknesses of traditional Apriori algorithm, an efficient analogous frequent item set mining algorithm was presented. With regard to the instance, association rule analysis was carried out for data of transformer defect in Shandong. Relevant results indicated that diverse attribute items were undoubtedly associated with each other to different degrees; in addition, the correlation obtained was adopted to perform operational maintenance for auxiliary equipment and parts, etc. that are vulnerable to defects.