The method of fault diagnosis based on NMF and SVM

Author

Yang Ying-hua ; Shan Ji-chang ; Chen Xiao-bo ; Qin Shu-kai

Author_Institution

Coll. of Inf. Sci. & Eng., Northeastern Univ., Shenyang

fYear

2008

fDate

2-4 July 2008

Firstpage

458

Lastpage

462

Abstract

As a new method of multivariate statistic analysis, non-negative matrix factorization (NMF) has been proposed for dealing with non-negative data. The results of NMF are non-negative and can be interpreted and understood easily, so they have specific physical meaning. In the production process, most data are non-negative. Due to this situation, a method of fault diagnosis based on NMF and SVM is presented in this paper. An on-line process monitoring model is built through NMF, and multi-fault classifiers are trained by SVM. The faults which are monitored by on-line monitoring model will be confirmed and identified by the fault classifiers. The simulation results of three water tank system show the effectiveness of this method.

Keywords

fault diagnosis; matrix decomposition; process monitoring; statistical analysis; tanks (containers); NMF; SVM; fault diagnosis; multifault classifiers; multivariate statistic analysis; nonnegative matrix factorization; online process monitoring model; Condition monitoring; Data engineering; Educational institutions; Fault diagnosis; Information science; Lagrangian functions; Principal component analysis; Statistical analysis; Support vector machine classification; Support vector machines; Fault diagnosis; Non-negative matrix factorization (NMF); Process monitoring; Support vector machine (SVM);

fLanguage

English

Publisher

ieee

Conference_Titel

Control and Decision Conference, 2008. CCDC 2008. Chinese

Conference_Location

Yantai, Shandong

Print_ISBN

978-1-4244-1733-9

Electronic_ISBN

978-1-4244-1734-6

Type

conf

DOI

10.1109/CCDC.2008.4597352

Filename

4597352