| 面向汽轮机振动故障诊断的FCM预处理方法 |
| |
| 引用本文： | 邴汉昆,丁常富,卢盛阳.面向汽轮机振动故障诊断的FCM预处理方法[J].电站系统工程,2012,28(4):13-15. |
| |
| 作者姓名： | 邴汉昆 丁常富 卢盛阳 |
| |
| 作者单位： | 1. 华北电力大学
2. 河北省电力研究院 |
| |
| 摘 要： | 提出一种故障诊断的方法,利用模糊C均值聚类,对汽轮机故障信号中提取的特征样本进行聚类分析,利用故障样本隶属度矩阵寻找其中一些故障特征不太明显甚至错误的样本,以此加强不同故障样本的特征,最后对优化的样本进行基于支持向量机训练,以此训练模型进行故障诊断。通过实验可以看出,经过优化后的样本,训练出的故障诊断模型精度得到了提高。
|
| 关 键 词： | 模糊C均值聚类 支持向量机 故障诊断 汽轮机振动 |
FCM Pretreatment Method for Fault Diagnosis of Turbine Vibration |
| |
| BING Han-Kun
,
DING Chang-Fu
,
LU Sheng-Yang.FCM Pretreatment Method for Fault Diagnosis of Turbine Vibration[J].Power System Engineering,2012,28(4):13-15. |
| |
| Authors: | BING Han-Kun DING Chang-Fu LU Sheng-Yang |
| |
| Affiliation: | BING Han-Kun,DING Chang-Fu,LU Sheng-Yang |
| |
| Abstract: | A fault diagnosis method is presented to classify the characteristics of the turbine fault signal using fuzzy C-means clustering in order to strengthen the different failure characteristics of the sample,find out some less obvious or even error features of the sample through membership matrix.Finally,training samples by SVM to get the model for fault diagnosis.As can be seen through experiments,the accuracy of fault diagnosis model trained by the optimized samples has been improved. |
| |
| Keywords: | fuzzy C-means clustering support vector machines fault diagnosis turbine vibration |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
