Fault Investigation of Rolling Element Bearing using Soft Computing
By: Agrawal, Pavan.
Contributor(s): Jayaswal, Pratesh.
Publisher: New Delhi STM Journals 2019Edition: Vol 6 (1) Jan- Apr.Description: 8-19p.Subject(s): Computer EngineeringOnline resources: Click Here In: Journal of artificial intelligence research and advances (JoAIRA)Summary: Rolling element bearing is the most significant mechanical component of any rotating machinery and used in vast applications from the surgical machine tool to aircraft etc. In the present research, an experimental investigation has been done using vibration analysis, continuous wavelet transform (CWT), and soft computing, i.e., support vector machine (SVM) and k-nearest neighbor (k-NN) for fault detection and diagnosis of rolling element bearing. Vibration signatures in the time domain are acquired from healthy and faulty bearings. Four complex-valued wavelets are considered. Out of four wavelets, mother wavelet is selected using relative wavelet energy (RWE) criterion and extracts the statistical features from the wavelet coefficient of raw vibration signatures. These statistical features are used as input of support vector machine and k-nearest neighbor for classification of faults. Finally, complex Morlet wavelet is selected and reports better results with support vector machine algorithm.Item type | Current location | Call number | Status | Date due | Barcode | Item holds |
---|---|---|---|---|---|---|
Articles Abstract Database | School of Engineering & Technology Archieval Section | Not for loan | 2021-2021463 |
Rolling element bearing is the most significant mechanical component of any rotating machinery and used in vast applications from the surgical machine tool to aircraft etc. In the present research, an experimental investigation has been done using vibration analysis, continuous wavelet transform (CWT), and soft computing, i.e., support vector machine (SVM) and k-nearest neighbor (k-NN) for fault detection and diagnosis of rolling element bearing. Vibration signatures in the time domain are acquired from healthy and faulty bearings. Four complex-valued wavelets are considered. Out of four wavelets, mother wavelet is selected using relative wavelet energy (RWE) criterion and extracts the statistical features from the wavelet coefficient of raw vibration signatures. These statistical features are used as input of support vector machine and k-nearest neighbor for classification of faults. Finally, complex Morlet wavelet is selected and reports better results with support vector machine algorithm.
There are no comments for this item.