A Feature Extraction Method Using Improved Multi-Scale Entropy for Rolling Bearing Fault Diagnosis

• Published in: Entropy (Basel, Switzerland) Vol. 20; no. 4; p. 212
• Main Authors: Ju, Bin, Zhang, Haijiao, Liu, Yongbin, Liu, Fang, Lu, Siliang, Dai, Zhijia
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 21-03-2018; MDPI
• Subjects: Bearing; bearing fault diagnosis; Entropy; Fault diagnosis; Feature extraction; Identification methods; improved multi-scale entropy; multi-scale entropy; Multiscale analysis; Similarity; Support vector machines
• ISSN: 1099-4300; 1099-4300
• DOI: 10.3390/e20040212

A feature extraction method named improved multi-scale entropy (IMSE) is proposed for rolling bearing fault diagnosis. This method could overcome information leakage in calculating the similarity of machinery systems, which is based on Pythagorean Theorem and similarity criterion. Features extracted from bearings under different conditions using IMSE are identified by the support vector machine (SVM) classifier. Experimental results show that the proposed method can extract the status information of the bearing. Compared with the multi-scale entropy (MSE) and sample entropy (SE) methods, the identification accuracy of the features extracted by IMSE is improved as well.