期刊名称：IOP Conference Series: Earth and Environmental Science
印刷版ISSN：1755-1307
电子版ISSN：1755-1315
出版年度：2019
卷号：252
期号：2
页码：1-6
DOI：10.1088/1755-1315/252/2/022111
出版社：IOP Publishing
摘要：Vibration signals of the rolling bearing are non-stationary, EEMD is occupied to decompose the signals; then ARMA model, which has higher accuracy, is established based on the signal principal components from the EEMD; then correlation dimension of the Auto-regressive (AR) parameters, taken from the ARMA model, serve as the feature vectors to be input to LSSVM for discriminating the actual conditions of rolling bearings. The results demonstrate that: the fault-patterns could be obtained accurately with the presented method, and it is an effective analysis method.
