Research on Familial Defect Recognition Method of Transformer Based on Correlation Analysis

As the key equipment of power system, the reliability of the transformer directly affects the safe operation of the power system. Due to the existence of familial defects in design, materials, and manufacturing processes, the probability of familial defects is relatively high, resulting in a significantly higher fault rate after commissioning. At present, the research on the recognition method of transformer's familial defects is mainly based on a single parameter or a small number of parameters. There is no comprehensive systematic method to automatically analyze and identify familial defects. Therefore, this paper proposes a familial defect recognition method based on correlation analysis. The method of identifying the transformer's familial defects based on correlation analysis mainly includes the following four parts. Firstly, collecting data such as ledger information, defect record information, manufacturer information, and years of operation of transformers and so on. Then, using statistical analysis, correlation analysis to comprehensively analyze the data. Further, from the analysis results, the features closely related to familial defects are discovered, and the association relationship between state quantity and familial defects is established. Finally, using statistical analysis, multi-dimensional analysis, hierarchical analysis, correlation analysis to study the recognition method of the familial defect for transformer and establish analysis model. By analyzing the association relationship between feature quantities and the familial defects of the transformer, this paper proposes a method to identify the familial defects of the transformer. At the same time, the analysis model of transformer familial defect is also established. It can be concluded from this paper that the method of recognizing the familial defects based on correlation analysis is effective and accurate. Applying it to field data, the familial defects can be recognized efficiently and accurately.