摘要
基于当前对电力变压器故障识别中的模式空间划分缺乏研究的情况 ,将模糊聚类技术引入了电力变压器油中溶解气体分析DGA。在对所收集的 183组电力变压器绝缘故障样本进行了多层树形聚类的基础上 ,通过多次分析绝缘故障 ,得到了一种准确程度较高的故障诊断方法。
Because of being short of studying how to carve up the patterns of insulation faults in the field of transformer fault diagnoses, fuzzy classification theory was applied to the problem of dissolved gas analysis(DGA)in transformer oil. 183 fault samples were partitioned into a hierarchy fault tree, and basing on which a new approach was presented for DGA, by which a insulation fault has to be diagnosed by stages. Through enhancing the veracity in every stages, a accurate method was attained.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2001年第2期37-41,共5页
Proceedings of the CSEE
关键词
变压器油
溶解气体分析
模糊模式
聚类
电力变压器
故障诊断
transformer DGA(dissolved gas analysis)
fuzzy model
multi layers classification
fault diagnose
