A comparative study of sound and vibration signals in detection of rotating machine faults using support vector machine and independent component analysis. (1st January 2014)

• Record Type: Journal Article
• Authors: Saimurugan, M.; Ramachandran, K.I.
• Is Part Of: International journal of data analysis techniques and strategies. Volume 6:Number 2(2014)
• Page Start: 188
• Page End: 204
• Publication Date: 2014-01-01
• Subjects: rotating machines -- fault diagnosis -- machine learning -- vibration -- sound -- support vector machine -- SVM -- independent component analysis -- ICA; Electronic data processing -- Periodicals; Database searching -- Periodicals; 005
• Languages: English
• ISSNs: 1755-8050

Abstract

In rotating machines, shaft and bearing are the critical components of interest for fault diagnosis. In recent years, the application of machine learning for fault diagnosis is gaining momentum. This paper presents the fault diagnosis of shaft and bearing using support vector machine (SVM). The experiments were conducted by simulating the 12 fault conditions of shaft and bearing. The statistical features were extracted from the collected vibration and sound signals and these features were given as an input to the classifier. The extracted statistical features were subjected to dimensionality reduction using independent component analysis (ICA) and classified using SVM. The obtained results of SVM and SVM with ICA are evaluated. The effectiveness of the vibration and sound signal for the fault diagnosis are discussed and compared.