改进型灰色关联算法在变压器故障诊断中的应用  

Application of improved grey relational algorithm in faults diagnosis of power transformers

作　　者：伍伟慧[1] 周媛媛[1] 周力行[1] 

机构地区：[1]长沙理工大学电气与信息工程学院,湖南长沙410004

出　　处：《电力科学与技术学报》2009年第3期45-48,53,共5页Journal of Electric Power Science And Technology

基　　金：湖南省自然科学基金(04JJ40034)

摘　　要：灰色关联分析已应用于电力变压器故障诊断,传统基于单一故障标准模式向量灰色关联分析算法的油中溶解气体分析(DGA)诊断模型精度有限.为此,提出一种改进型灰色关联算法,该算法在充分考虑DGA数据分散性的基础上,将每类故障的标准故障模式向量由原来算法中的1个扩充到6个,并给出每类故障的DGA数据分布范围,增大诊断信息量;利用关联分析原理,求出待诊模式与各类故障标准模式的灰色关联度,得到故障诊断判定.实例分析证明,所提算法的诊断准确率高于原来的普通灰色关联方法.Grey relational analysis （GRA） has applied in power transformer faults diagnose. Traditional Dissolved Gas Analysis （DGA） diagnostic model based on the GRA algorithm of single fault standard model vector has limited precision. An improved GRA algorithm is presented in this paper. The dispersion of DGA data is considered ; every kind of fault standard model vector is extended to six vectors ; the DGA data distribution range for every kind of fault is given; and the diagnose information is thus increased; the grey correlation degree of diagnostic model and all kinds of fault standard models are obtained to realize the fault diagnosis by using grey relational analysis theory. Example analysis results prove that the diagnostic accuracy of the presented algorithm is much higher than that of the others.

关 键 词：变压器 改进型灰色关联 油中溶解气体分析 故障诊断 

分 类 号：TM401.1[电气工程—电器]