Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Chinese Journal of Sensors and Actuators》 2006-01
Add to Favorite Get Latest Update

Research of Machine Fault Diagnosis Based on PCA and SVDD

PAN Ming-qing1, ZHOU Xiao-jun1, WU Rui-ming2, LEI Liang-yu3 (1.Institute of Modern Manufacture Engineering, Zhejiang University, Hangzhou 310027,China; 2.Dept. Machinery Engineering Zhejinang Unversity of Science and Technology,Hangzhou 310012,China; 3.Dept. Machinery Engineering Jiangsu Teachers University of Technology,Jiangsu Changzhou 213001,China)  
An Oneclass classification method-support vector data description(SVDD) is proposed for machine fault diagnosis, in this condition fault samples are always scarce. This method can build up one-class classification to distinguish normal and abnormal condition only using normal samples. In the process of test, principal component analysis is used as data preprocessing to extract the feature index from vibration signal statistic features as the input of SVDD classifier. The test result shows that after the extraction of PCA, the SVDD classifier distinguished the normal and fault condition finely, and it also has good recognized ability to unknown fault samples.
【CateGory Index】: TH17
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved