|
| 摘要: |
| 电力变压器油中溶解气体分析DGA(Dissolved Gas Analysis)是电力变压器故障诊断的重要方法,为了克服模糊C-均值算法存在的聚类中心数不容易确定,以及容易陷入极小的问题,在运用改进遗传算法的基础上,探讨了一种将自适应遗传算法和模糊C-均值相结合的遗传-模糊聚类算法,并将其应用于DGA电力变压器数据分析,实现了变压器的故障诊断。实验数据表明:该算法收敛速度快,能有效地对样本进行聚类,提高了识别故障率。 |
| 关键词: 电力变压器,溶解气体分析,模糊C-均值聚类算法,遗传算法,故障诊断 |
| DOI： |
| 分类号:TM411 TM835 |
| 基金项目:中国博士后科学基金项目(2006040933)~~ |
|
| Dissolved gas analysis based on GA and FCM for power transformer fault diagnosis |
|
TIAN Zhiguang ZHANG Huifen
|
| Abstract: |
| DGA(Dissolved Gas Analysis) is a method of power transformer fault diagnosis.Because the clustering center of FCM(Fuzzy C-Mean clustering algorithm) is difficult to determine and the local minimum occurs easily,a method,which combines the adaptive GA(Genetic Algorithm) and FCM together,is applied to DGA data analysis.The experimental data shows that,the proposed algorithm converges quickly,clusters samples effectively and recognizes faults properly. |
| Key words: power transformer,dissolved gas analysis,fuzzy C-mean clustering,genetic algorithm,fault diagnosis |
