基于主成分分析和遗传程序设计的油浸式变压器故障诊断  

Oil-immersed transformer fault diagnosis based on PCA and GP

作　　者：董卓[1] 朱永利[1] 张敬敬[1] 陈哲星[1] 胡资斌[1] 

机构地区：[1]华北电力大学电子与电气工程学院,河北保定071003

出　　处：《河北工业科技》2010年第6期477-479,488,共4页Hebei Journal of Industrial Science and Technology

摘　　要：提出了一种基于主成分分析和遗传程序设计的分类模型,并应用于油浸式变压器故障诊断中。该分类模型以变压器油中溶解气体的DGA数据为评价指标;通过对所采集数据进行主成分分析,减少了数据的冗余度,从而大大降低了问题的复杂度。实验证明该模型是高效可行的。We proposed a classification model based on principal component analysis and genetic programming, and applied it to fault diagnosis of oil-immersed transformers. The classification model used the DGA data of dissolved gas in the transformer oil as the evaluation indicalors; It reduced data redundancy by using the PCA to analyse the collected data, which greatly reduces the complexity of the problem. Experimental results show that the model is efficient and feasible.

关 键 词：油浸式变压器 主成分分析 遗传程序设计 故障诊断 

分 类 号：TM411[电气工程—电器]