0
Research Papers

A Joint Kurtosis-Based Adaptive Bandstop Filtering and Iterative Autocorrelation Approach to Bearing Fault Detection

[+] Author and Article Information
Ming Liang

e-mail: liang@eng.uOttawa.ca
Department of Mechanical Engineering,
University of Ottawa,
Ottawa K1N 6N5, Canada

Chuan Li

Department of Mechanical Engineering,
University of Ottawa,
Ottawa K1N 6N5, Canada;
Engineering Laboratory for Detection,
Control and Integrated System,
Chongqing Technology and Business University,
Chongqing 400067, China

Shumin Hou

Department of Mechanical Engineering,
University of Ottawa,
Ottawa K1N 6N5, Canada

1Corresponding author.

Contributed by the Design Engineering Division of ASME for publication in the JOURNAL OF VIBRATION AND ACOUSTICS. Manuscript received October 6, 2012; final manuscript received May 5, 2013; published online June 18, 2013. Assoc. Editor: Patrick S. Keogh.

J. Vib. Acoust 135(5), 051026 (Jun 18, 2013) (17 pages) Paper No: VIB-12-1282; doi: 10.1115/1.4024610 History: Received October 06, 2012; Revised May 05, 2013

This paper reports a bearing fault detection method based on kurtosis-based adaptive bandstop filtering (KABS) and iterative autocorrelation (IAC). The interferences in the bearing signal can be removed by KABS filtering, whereas IAC is employed for noise reduction and signal enhancement. In the KABS method, two window-merging schemes are proposed to identify the frequency bands potentially containing interferences and to preserve those covering fault frequencies. Issues related to the selection of the number of autocorrection iterations are also discussed. The proposed method can be used for bearing fault detection in a low signal-to-noise ratio (SNR) and low signal-to-interference ratio (SIR) environment. The implementation of the proposed method does not require prior knowledge of the fault-excited resonant frequency. The performance of the proposed method has been examined by simulation analysis, with favorable comparisons to the Hilbert enveloping, energy operator, and spectrum kurtosis methods. Its effectiveness in bearing fault detection has also been demonstrated using experimental data.

Copyright © 2013 by ASME
Your Session has timed out. Please sign back in to continue.

References

McInerny, S. A., and Dai, Y., 2003, “Basic Vibration Signal Processing for Bearing Fault Detection,” IEEE Trans. Educ., 46, pp. 149–156. [CrossRef]
Bozchalooi, I. S., and Liang, M., 2008, “A Joint Resonance Frequency Estimation and In-Band Noise Reduction Method for Enhancing the Detectability of Bearing Fault Signals,” Mech. Syst. Signal Process., 22, pp. 915–933. [CrossRef]
Bozchalooi.I. S., and Liang, M., 2009, “Parameter-Free Bearing Fault Detection Based on Maximum Likelihood Estimation and Differentiation,” Meas. Sci. Technol., 20, p. 065102. [CrossRef]
Liang, M., and Bozchalooi, I. S., 2010, “An Energy Operator Approach to Joint Application of Amplitude and Frequency-Demodulations for Bearing Fault Detection,” Mech. Syst. Signal Process., 24, pp. 1473–1494. [CrossRef]
Lin, J., and Zuo, M., 2003, “Gearbox Fault Diagnosis Using Adaptive Wavelet Filter,” Mech. Syst. Signal Process., 17, pp. 1259–1269. [CrossRef]
Qiu, H., Lee, J., Lin, J., and Yu, G., 2006, “Wavelet Filter-Based Weak Signature Detection Method and Its Application on Rolling Element Bearing Prognosis,” J. Sound Vib., 289, pp. 1066–1090. [CrossRef]
Bozchalooi, I. S., and Liang, M., 2007, “A Smoothness Index-Guided Approach to Wavelet Parameter Selection in Signal De-Noising and Fault Detection,” J. Sound Vib., 308, pp. 246–267. [CrossRef]
Immovilli, F., Cocconcelli, M., Bellini, A., and Rubini, R., 2009, “Detection of Generalized-Roughness Bearing Fault by Spectral-Kurtosis Energy of Vibration or Current Signals,” IEEE Trans. Ind. Electron., 56, pp. 4710–4717. [CrossRef]
Combet, F., and Gelman, L., 2009, “Optimal Filtering of Gear Signals for Early Damage Detection Based on the Spectral Kurtosis,” Mech. Syst. Signal Process., 23, pp. 652–668. [CrossRef]
Antoni, J., 2006, “The Spectral Kurtosis: A Useful Tool for Characterizing Non-Stationary Signals,” Mech. Syst. Signal Process., 20, pp. 282–307. [CrossRef]
Wang, Y. X., and Liang, M., 2010, “An Adaptive SK Technique and Its Application for Fault Detection of Rolling Element Bearings,” Mech. Syst. Signal Process., 25, pp. 1750–1764. [CrossRef]
Donoho, D. L., 1995, “De-Noising by Soft-Thresholding,” IEEE Trans. Inf. Theory, 41, pp. 613–627. [CrossRef]
Donoho, D. L., and Johnstone, I. M., 1994, “Ideal Spatial Adaptation Via Wavelet Shrinkage,” Biometrika, 81, pp. 425–455. [CrossRef]
Qiu, H., Lee, J., Lin, J., and Yu, G., 2003, “Robust Performance Degradation Assessment Methods for Enhanced Rolling Element Bearing Prognostics,” Adv. Eng. Inf., 17, pp. 127–140. [CrossRef]
Lin, J., Zuo, M., and Fyfe, K., 2004, “Mechanical Fault Detection Based on the Wavelet De-Noising Technique,” ASME J. Vibr. Acoust., 126, pp. 9–16. [CrossRef]
Rafieea, J., and Tse, P. W., 2009, “Use of Autocorrelation of Wavelet Coefficients for Fault Diagnosis,” Mech. Syst. Signal Process., 23, pp. 1554–1572. [CrossRef]
Hong, H. B., 2007, “Rotating Machinery Monitoring: Feature Extraction, Signal Separation, and Fault Severity Evaluation,” Ph.D. thesis, University of Ottawa, Ottawa, Canada.
Wang, Y., Jin, L., Yao, M., and Yin, Q., 1999, “Research on Performance of Suppressing Noise by Cyclic Auto-Correlation Function,” J. Data Acquis. Process., 14, pp. 148–152. [CrossRef]
Li, Y. B., Yue, X., and Yang, X. Y., 2004, “Estimation of Sinusoidal Parameters in Powerful Noise by Multi-Layer Autocorrelation,” J. Harbin Eng. Univ., 25, pp. 525–528. [CrossRef]
Zheng, Z. W., and Zhang, F., 2010, “Comparative Study on Weak Signal Detection Algorithms,” International Conference on E-Business and E-Government (ICEE2010), Guangzhou, China, May 7–9, pp. 3825–3828. [CrossRef]
Rudoy, D., Basu, P., Quatieri, T. F., Dunn, B., and Wolfe, P. J., 2008, “Adaptive Short-Time Analysis-Synthesis for Speech Enhancement,” IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP2008), Las Vegas, NV, March 31–April 4, pp. 4905–4908. [CrossRef]
Pachaud, C., Salvetat, R., and Fray, C., 1997, “Crest Factor and Kurtosis Contributions to Identify Defects Inducing Periodical Impulsive Forces,” Mech. Syst. Signal Process., 11, pp. 903–916. [CrossRef]
Morgenstern, J., 1973, “Note on a Lower Bound of the Linear Complexity of the Fast Fourier Transform,” J. ACM, 20, pp. 305–306. [CrossRef]
Press, W. H., Teukolsky, S. A., Vetterling, W. T., and Flannery, B. P., 1992, Numerical Recipes in C: The Art of Scientific Computing, 2nd ed., Cambridge University, Cambridge, UK, pp. 545–546.

Figures

Grahic Jump Location
Fig. 1

Illustration of the kurtosis-based narrow band filtering. The solid line indicates the spectra of two interference components; the trapezoids are the narrow-band filter banks with the kurtosis value of the filtered signal shown inside each of them.

Grahic Jump Location
Fig. 2

Illustration of impulsive fault signal appearing as discrete spectral lines. (a) Simulated time domain signal and (b) frequency domain representation.

Grahic Jump Location
Fig. 3

Illustration of the bandstop filtering method. (a) Simulated time domain signal mixed with two interference components (120 Hz and 1100 Hz), (b) frequency domain representation of the signal in (a), (c) the ideal bandstop filter to remove interferences, (d) frequency domain representation of the filtered signal, and (e) time domain representation of the filtered signal.

Grahic Jump Location
Fig. 4

State transition diagram of frequency window merging process

Grahic Jump Location
Fig. 5

Block diagram of the interference removal process

Grahic Jump Location
Fig. 6

Illustration of RAR overlap of an impulsive signal as the number of IAC iterations increases: (a) original pulse signal, (b) one-iteration IAC, (c) two-iteration IAC, (d) three-iteration IAC, (e) four-iteration IAC, and (f) five-iteration IAC

Grahic Jump Location
Fig. 7

Block diagram of the proposed bearing fault detection method

Grahic Jump Location
Fig. 8

The simulated fault impulsive signal with different SNRs: (a) the signal without interference, (b) the signal with interferences of low frequencies (5, 20, 60, and 120 Hz), (c) the signal with interferences of high frequencies (650, 1100, 1900, and 2300 Hz), and (d) the signal with multiband interferences of both low frequencies (5, 20, 60, and 120 Hz) and high frequencies (650, 1100, 1900, and 2300 Hz)

Grahic Jump Location
Fig. 9

The kurtosis values of the simulated signals with different SNRs: (a) the signal without interference, (b) the signal with interferences of low frequencies (5, 20, 60, and 120 Hz), (c) the signal with interferences of high frequencies (650, 1100, 1900, and 2300 Hz), and (d) the signal with multiband interferences of both low frequencies (5, 20, 60, and 120 Hz) and high frequencies (650, 1100, 1900, and 2300 Hz)

Grahic Jump Location
Fig. 10

The kurtosis values and the resulting bandstop filters (SNR = −20): (a) the kurtosis value of signal with low frequencies (5, 20, 60, and 120 Hz), (b) the resultant bandstop filter of (a), (c) the kurtosis value of signal with high frequencies (650, 1100, 1900, and 2300 Hz), (d) the resultant bandstop filter of (c), (e) the kurtosis value of signal with multiband frequencies (5, 20, 60, 120, 650, 1100, 1900, and 2300 Hz), and (f) the resultant bandstop filter of (e)

Grahic Jump Location
Fig. 11

Fault detection results for the simulated signal (without interference) obtained by (a) Hilbert enveloping, (b) EO method, (c) SK method, (d) KABS without IAC, (e) KABS plus one-iteration IAC, and (f) KABS plus two-iteration IAC

Grahic Jump Location
Fig. 12

Fault diagnosis results for the simulated signals (with low frequency interferences) obtained by (a) Hilbert enveloping method, (b) EO method, (c) SK method, (d) KABS without IAC, (e) KABS plus one-iteration IAC, and (f) KABS plus two-iteration IAC

Grahic Jump Location
Fig. 13

Fault detection results for the simulated signals (with high frequency interferences) obtained by (a) Hilbert enveloping method, (b) EO method, (c) SK method, (d) KABS without IAC, (e) KABS plus one-iteration IAC, and (f) KABS plus two-iteration IAC

Grahic Jump Location
Fig. 14

Fault detection results for the simulated signals (with multiband interferences) obtained by (a) Hilbert enveloping method, (b) EO method, (c) SK method, (d) KABS without IAC, (e) KABS plus one-iteration IAC, and (f) KABS plus two-iteration IAC

Grahic Jump Location
Fig. 15

Experimental setup for bearing fault detection

Grahic Jump Location
Fig. 16

Bearing outer race fault at different positions: (a) top position, (b) middle (side) position, (c) bottom position

Grahic Jump Location
Fig. 17

Detection of top-positioned outer race fault using classical methods: (a) original signal, (b) frequency spectrum of the signal, and (c) envelope spectrum

Grahic Jump Location
Fig. 18

Detection of middle (side)-positioned outer race fault using classical methods: (a) original signal, (b) frequency spectrum of the signal, and (c) envelope spectrum

Grahic Jump Location
Fig. 19

Detection of middle (side)-positioned outer race fault using the proposed method: (a) bandstop filtered signal, (b) frequency spectrum of the signal in (a), and (c) envelope spectrum of (a)

Grahic Jump Location
Fig. 20

Detection of the bottom-positioned outer race fault based on classical method: (a) original signal, (b) frequency spectrum of the signal, and (c) envelope spectrum

Grahic Jump Location
Fig. 21

Detection of the bottom-positioned outer race fault using the KABS + IAC method: (a) bandstop filtered signal, (b) frequency spectrum of the signal in (a), (c) envelope spectrum of the signal in (a), (d) envelope spectrum of KABS + one-iteration IAC, and (e) envelope spectrum of KABS + two-iteration IAC

Grahic Jump Location
Fig. 22

Detection of inner race fault based on classical methods: (a) original signal, (b) frequency spectrum of (a), and (c) envelope spectrum

Grahic Jump Location
Fig. 23

Detection of inner race fault using the KABS + IAC method: (a) bandstop filtered signal, (b) frequency spectrum of (a), (c) envelope spectrum of (a), (d) envelope spectrum of KABS + one-iteration IAC, and (e) envelope spectrum of KABS + two-iteration IAC

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