0
Research Papers

Incipient Bearing Fault Feature Extraction Based on Minimum Entropy Deconvolution and K-Singular Value Decomposition

[+] Author and Article Information
Guangming Dong

State Key Laboratory of Mechanical
Systems and Vibration,
Shanghai Jiao Tong University,
Shanghai 200240, China
e-mail: gmdong@sjtu.edu.cn

Jin Chen

State Key Laboratory of Mechanical
Systems and Vibration,
Shanghai Jiao Tong University,
Shanghai 200240, China

Fagang Zhao

Shanghai Institute of Satellite Engineering,
251 Huaning Road,
Minhang District,
Shanghai 200240, China

Manuscript received January 13, 2017; final manuscript received June 29, 2017; published online August 24, 2017. Assoc. Editor: Ivan Selesnick.

J. Manuf. Sci. Eng 139(10), 101006 (Aug 24, 2017) (12 pages) Paper No: MANU-17-1020; doi: 10.1115/1.4037419 History: Received January 13, 2017; Revised June 29, 2017

Machinery condition monitoring and fault diagnosis are essential for early detection of equipment malfunctions or failures, which insure productivity, quality, and safety in the manufacturing process. This paper aims at extracting fault features of rolling element bearings at the incipient fault stage. K-singular value decomposition (K-SVD), one technique for sparse representation of signals, is used for study. In K-SVD, its dictionary is trained from data by machine learning techniques, which allows more flexibility to adapt to variation of real signals than the predefined dictionaries. Analysis on simulated bearing signals and real signals shows that K-SVD can give better bearing fault features than the predefined dictionaries such as wavelet dictionaries. However, during our simulation study, K-SVD was found to have large representation error under heavy noise. To reduce the noise effect, minimum entropy deconvolution (MED) is used as a prefilter. The combination of MED and K-SVD is proposed for incipient bearing fault detection. The method is verified by simulation and experimental study. It is shown that the proposed method can effectively extract the impulsive fault feature of the tested bearing at its incipient fault stage.

Copyright © 2017 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

The process of dictionary learning

Grahic Jump Location
Fig. 2

Input, output, and objective of K-SVD

Grahic Jump Location
Fig. 3

Flowchart of the K-SVD algorithm

Grahic Jump Location
Fig. 4

The simulation bearing signal with inner race fault

Grahic Jump Location
Fig. 5

Nine randomly selected atoms from the learned dictionary

Grahic Jump Location
Fig. 6

The denoised signal using K-SVD method

Grahic Jump Location
Fig. 7

The effect of noise on reconstruction accuracy

Grahic Jump Location
Fig. 8

The effect of total number of atoms on K-SVD: (a) reconstruction accuracy and (b) computation time

Grahic Jump Location
Fig. 9

The comparison of approximation error between K-SVD dictionary and predefined dictionaries

Grahic Jump Location
Fig. 10

Bearing fault feature extraction based on MED and K-SVD

Grahic Jump Location
Fig. 11

The original simulated signal: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 12

The denoised signal using MED: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 13

The denoised signal using MED and K-SVD: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 14

The denoised signal using K-SVD only: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 15

The denoised signal using SK filter: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 16

Testing object: (a) locations of acceleration sensors and (b) the sketch of sensors installations

Grahic Jump Location
Fig. 17

Experiment equipment

Grahic Jump Location
Fig. 18

Failure on rolling element bearing

Grahic Jump Location
Fig. 19

The root-mean-square (RMS) value during the whole life of the tested bearing

Grahic Jump Location
Fig. 20

The time waveform and its frequency spectrum in each stage: (a) and (b) normal stage; (c) and (d) incipient fault stage; (e) and (f) severe fault stage

Grahic Jump Location
Fig. 21

The original signal at 1877th min: (a) time waveform and (b) frequency spectrum

Grahic Jump Location
Fig. 22

The filtered signal using MED: (a) time waveform and (b) frequency spectrum

Grahic Jump Location
Fig. 23

Nine atoms in K-SVD dictionary

Grahic Jump Location
Fig. 24

The denoised signal using MED and K-SVD: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 25

The denoised signal using SK filter: (a) time waveform and (b) envelope frequency spectrum

Grahic Jump Location
Fig. 26

Analysis results using predefined dictionary Db8: (a) denoised signal using MP, (b) envelope spectrum of (a), (c) denoised signal using BP, and (d) envelope spectrum of (c)

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In