引用本文: | 刘卫华,廖瑞金,杨丽君.基于点密度加权核模糊聚类的变压器故障诊断方法[J].电力自动化设备,2012,32(6): |
| LIU Weihua,LIAO Ruijin,YANG Lijun.Power transformer fault diagnosis based on dot density weighted fuzzy kernel clustering[J].Electric Power Automation Equipment,2012,32(6): |
|
摘要: |
变压器油中溶解气体分析(DGA)是电力变压器绝缘诊断的重要方法。针对模糊C- 均值聚类算法(FCM)用于DGA时存在可分性差和等趋势划分等问题,用样本点分布密度大小作为权值,结合核函数的增强可分性,提出点密度加权模糊核C- 均值聚类算法,并将其用于变压器DGA 数据分析,从而实现变压器的故障诊断。实例分析结果表明该算法能快速、有效地对样本进行聚类,且特别适用于含有噪声样本的环境。 |
关键词: 点密度 核函数 FCM 变压器 DGA 故障诊断 模糊理论 聚类算法 |
DOI: |
分类号: |
基金项目: |
|
Power transformer fault diagnosis based on dot density weighted fuzzy kernel clustering |
LIU Weihua, LIAO Ruijin, YANG Lijun
|
State Key Laboratory of Transmission & Distribution Equipment and Power System Security and New Technology,Chongqing University,Chongqing 400044,China
|
Abstract: |
DGA(Dissolved Gas Analysis) is an important method of power transformer insulation diagnosis. There are defects in the fuzzy C-means clustering algorithm for DGA,such as poor separability and difficult isoclinal dividing,for which,the dot density weighted kernel fuzzy C-means clustering algorithm is proposed. It takes the dot density as the weight and combines with the enhanced separability of kernel function in DGA for transformer fault diagnosis. Case study shows that the proposed algorithm can quickly and effectively carry out sample clustering,suitable especially for the samples with noise. |
Key words: dot density kernel function FCM electric transformers DGA fault diagnosis fuzzy theory clustering algorithms |