| 基于支持向量机的涡轮泵故障诊断方法研究 |
| |
| 引用本文： | 田路,田干,张炜,李亮. 基于支持向量机的涡轮泵故障诊断方法研究[J]. 控制工程, 2007, 0(Z2) |
| |
| 作者姓名： | 田路 田干 张炜 李亮 |
| |
| 作者单位： | 第二炮兵工程学院二系 陕西西安710025 |
| |
| 摘 要： | 针对涡轮泵的几种典型常见故障,应用支持向量机(SVM)构造多元分类器,解决故障诊断等多分类问题,在此基础上建立了基于支持向量机的故障诊断模型。试验结果表明,与神经网络相比,采用支持向量机进行故障诊断可得到更高的精度,表现出较好的分类和抗噪能力,而且效率高于神经网络,说明该方法是有效、可行的。
|
| 关 键 词： | 涡轮泵 支持向量机 故障诊断 多分类 |
On Turbo Pump Fault Diagnosis Based on Support Vector Machine |
| |
| TIAN Lu,TIAN Gan,ZHANG Wei,LI Liang. On Turbo Pump Fault Diagnosis Based on Support Vector Machine[J]. Control Engineering of China, 2007, 0(Z2) |
| |
| Authors: | TIAN Lu TIAN Gan ZHANG Wei LI Liang |
| |
| Abstract: | To the frequent fault types of turbo pump,the multi-classification machine is constructed applying SVM to solve the multi-classficatiom problem such as fault diagnosis.So fault diagnosis method of turbo pump based on SVM is proposed.The research shows that the method can achieve more excellent diagnosis accuracy,has higher classification and anti-noise ability and shorter training time than the methods based on artificial neural networks.The results show that the method is efficient and feasible,especially for small samples. |
| |
| Keywords: | turbo pump support vector machine fault diagnosis multi-classification |
| 本文献已被 CNKI 等数据库收录！ |
|
