| PCA特征抽取与SVM多类分类在传感器故障诊断中的应用 |
| |
| 引用本文： | 彭红星,陈祥光,徐巍.PCA特征抽取与SVM多类分类在传感器故障诊断中的应用[J].数据采集与处理,2010,25(1). |
| |
| 作者姓名： | 彭红星 陈祥光 徐巍 |
| |
| 作者单位： | 1. 北京理工大学化工与环境学院,北京,100081;河南理工大学计算机科学与技术学院,焦作,454159
2. 北京理工大学化工与环境学院,北京,100081 |
| |
| 基金项目： | 北京理工大学校基础研究基金 |
| |
| 摘 要： | 为了提高传感器故障诊断的准确率,提出了基于主元分析(PCA)特征抽取和支持向量机(SVM)多类分类的故障诊断方法.该方法通过对传感器输出信号进行小波包分解产生原始特征数据,然后采用PCA特征抽取得到二次特征向量,增强传感器各个状态模式的可分性.二次特征输入到二叉树SVM多类分类算法设计的分类器实现传感器故障诊断.仿真实验结果表明,这种结合了PCA特征抽取和SVM分类的诊断方法准确率高,其诊断效果优于直接采用原始特征进行分类的情况.
|
| 关 键 词： | 传感器故障诊断 小波包分解 主元分析 特征抽取 支持向量机 |
Application of PCA Feature Extraction and SVM Multi-classification on Sensor Fault Diagnosis |
| |
| Peng Hongxing,Chen Xiangguang,Xu Wei.Application of PCA Feature Extraction and SVM Multi-classification on Sensor Fault Diagnosis[J].Journal of Data Acquisition & Processing,2010,25(1). |
| |
| Authors: | Peng Hongxing Chen Xiangguang Xu Wei |
| |
| Affiliation: | Peng Hongxing1,2,Chen Xiangguang1,Xu Wei1(1.School of Chemical Engineering & Environment,Beijing Institute of Technology,Beijing,100081,China,2.School of Computer Science & Technology,Henan Polytechnic University,Jiaozuo,454159,China) |
| |
| Abstract: | To improve the accuracy of the sensor fault diagnosis,a novel sensor fault diagnosis approach is presented based on principle component analysis(PCA) feature extraction and support vector machines(SVM) multi-classification.Original features are acquired by the wavelet packet decomposition of the sensor output signal and processed by PCA to produce further features.The separability of various sensor operation patterns can be enhanced.With these further features,the binary tree SVM multi-classification classi... |
| |
| Keywords: | sensor fault diagnosis wavelet packet decomposition principle component analysis feature extraction support vector machines |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
