Title :
Improved Kernel PCA Based on Wavelet for Fault Detection
Author :
Wu, Hongyan ; Huang, Daoping
Author_Institution :
Coll. of Autom. Sci. & Technol., South China Univ. of Technol., Guangzhou
Abstract :
KPCA is a promising method for solving nonlinear system in chemical process fault monitoring. In this paper, in order to improve the accuracy of KPCA for fault detection, a new method combined with wavelet is developed. Simulation results are given to show that the proposed approach has superior to KPCA in process monitoring performance.
Keywords :
chemical industry; fault diagnosis; principal component analysis; process monitoring; statistical process control; wavelet transforms; chemical process fault monitoring; fault detection; kernel principal component analysis; multivariate statistical control; nonlinear system solving; wavelet transform; Chemical technology; Educational institutions; Eigenvalues and eigenfunctions; Electromagnetic interference; Fault detection; Kernel; Monitoring; Principal component analysis; Wavelet analysis; Wavelet transforms;
Conference_Titel :
Computing, Communication, Control, and Management, 2008. CCCM '08. ISECS International Colloquium on
Conference_Location :
Guangzhou
Print_ISBN :
978-0-7695-3290-5
DOI :
10.1109/CCCM.2008.26
