SVD and Hankel matrix based de-noising approach for ball bearing fault detection and its assessment using artificial faults. (March 2016)

• Record Type: Journal Article
• Authors: Golafshan, Reza; Yuce Sanliturk, Kenan
• Is Part Of: Mechanical systems and signal processing. Volume 70/71(2016)
• Page Start: 36
• Page End: 50
• Publication Date: 2016-03
• Subjects: Ball bearing fault detection -- De-noising -- Singular value decomposition -- Time domain analysis -- Frequency domain analysis; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.08.012
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: Ball bearings remain one of the most crucial components in industrial machines and due to their critical role, it is of great importance to monitor their conditions under operation. However, due to the background noise in acquired signals, it is not always possible to identify probable faults. This incapability in identifying the faults makes the de-noising process one of the most essential steps in the field of Condition Monitoring (CM) and fault detection. In the present study, Singular Value Decomposition (SVD) and Hankel matrix based de-noising process is successfully applied to the ball bearing time domain vibration signals as well as to their spectrums for the elimination of the background noise and the improvement the reliability of the fault detection process. The test cases conducted using experimental as well as the simulated vibration signals demonstrate the effectiveness of the proposed de-noising approach for the ball bearing fault detection. Highlights: SVD based de-noising method is applied to bearings vibration signals for diagnosis. Hankel matrix is used in SVD process both for time and frequency domains signals. De-noised signals are reconstructed via the modified Hankel matrices. De-noised signals provide better fault detection capability.