Abstract
For automatic detection and diagnosis of localized defects in rolling element bearings, bicoherence spectra are used to derive features that signify the condition of a bearing. These features quantitatively describe the degree of phase correlation among any three harmonics of bearing characteristic defect frequencies. Employing these features, a linear discriminant classifier is implemented to detect localized defects on a roller and the outer race of a bearing. Experimental results show that the proposed scheme is effective in bearing defect detection and sensitive to incipient defects.
Keywords
Get full access to this article
View all access options for this article.
