灰色关联分析在变压器DGA缺码数据中的应用研究

Application Research of Grey Relational Analysis in Transformer DGA Code Absence Data

摘要: 三比值法是电力变压器进行潜伏性故障诊断的有效方法之一，但该方法存在缺码问题。本文根据收集到的DGA数据，发现相同故障样本中特征气体的增减变化趋势相同，不同故障特征气体变化趋势有明显不同。本文通过挖掘DGA中各数据的变化趋势，运用数据灰色关联度理论，使用基于灰色关联度的DGA故障识别方法，对DGA缺码故障变压器进行故障诊断。弥补了三比值法的不足，提高变压器故障诊断精度，研究成果具有重要的理论价值和工程实用价值，可以大力推广，给电力系统及社会带来较大的经济效益。

Abstract: Three-ratio method is one of the effective methods for latent fault diagnosis of power transfor-mers, but there is a lack of code in this method. According to the DGA data collected in this paper, it is found that the variation trend of the characteristic gas in the same fault samples is the same, and the variation trend of different fault features is obviously different. In this paper, by excavate the changing trend of each data in DGA, using the grey relational theory and using the DGA fault identification method based on grey relational analysis, the fault diagnosis of code absence power transformers in DGA is carried out. It makes up for the deficiency of the three ratio method and improves the accuracy of transformer fault diagnosis. The research results have important theoretical value and practical value of engineering. It can be popularized vigorously and bring great economic benefits to the power system and the society.