Title :
Roller bearings fault diagnosis based on LS-SVM
Author :
Sui, Wentao ; Zhang, Dan ; Wang, Wilson
Author_Institution :
Sch. of Mech. Eng., Shandong Univ. of Technol., Zibo, China
Abstract :
A new method of roller bearings fault diagnosis based on least squares support vector machines (LS-SVM) was presented. Feature selection method based on simulated annealing (SA) algorithm was discussed in this paper. LS-SVM classifier was constructed for bearing faults. Compared with the Artificial Neural Network based method, the LS-SVM based method possessed desirable advantages. Experiment shows that the presented method is able to reliably recognize different fault categories.
Keywords :
fault diagnosis; feature extraction; least squares approximations; mechanical engineering computing; pattern classification; rolling bearings; simulated annealing; support vector machines; LS-SVM classifier; feature selection method; least squares support vector machine; roller bearing fault diagnosis; simulated annealing algorithm; Artificial intelligence; Artificial neural networks; Chemical industry; Fault diagnosis; Least squares methods; Risk management; Rolling bearings; Signal processing algorithms; Support vector machine classification; Support vector machines; fault diagnosis; feature evaluation; signal processing; support vector machine;
Conference_Titel :
Automation and Logistics, 2009. ICAL '09. IEEE International Conference on
Conference_Location :
Shenyang
Print_ISBN :
978-1-4244-4794-7
Electronic_ISBN :
978-1-4244-4795-4
DOI :
10.1109/ICAL.2009.5262645
