Low Speed Bearing Fault Diagnosis Based on EMD-CIIT Histogram Entropy and KFCM Clustering

doi:10.1007/s12204-019-2108-0

Abstract

Abstract: In view of weak defect signals and large acoustic emission (AE) data in low speed bearing condition monitoring, we propose a bearing fault diagnosis technique based on a combination of empirical mode decomposition (EMD), clear iterative interval threshold (CIIT) and the kernel-based fuzzy c-means (KFCM) eigenvalue extraction. In this technique, we use EMD-CIIT and EMD to complete the noise removal and to extract the intrinsic mode functions (IMFs). Then we select the first three IMFs and calculate their histogram entropies as the main fault features. These features are used for bearing fault classification using KFCM technique. The result shows that the combined EMD-CIIT and KFCM algorithm can accurately identify various bearing faults based on AE signals acquired from a low speed bearing test rig.

Key words: empirical mode decomposition - clear iterative interval threshold (EMD-CIIT)| kernel-based fuzzy c-means (KFCM)| acoustic emission (AE) signals| low speed machine| roller element bearing

摘要： In view of weak defect signals and large acoustic emission (AE) data in low speed bearing condition monitoring, we propose a bearing fault diagnosis technique based on a combination of empirical mode decomposition (EMD), clear iterative interval threshold (CIIT) and the kernel-based fuzzy c-means (KFCM) eigenvalue extraction. In this technique, we use EMD-CIIT and EMD to complete the noise removal and to extract the intrinsic mode functions (IMFs). Then we select the first three IMFs and calculate their histogram entropies as the main fault features. These features are used for bearing fault classification using KFCM technique. The result shows that the combined EMD-CIIT and KFCM algorithm can accurately identify various bearing faults based on AE signals acquired from a low speed bearing test rig.

关键词: empirical mode decomposition - clear iterative interval threshold (EMD-CIIT)| kernel-based fuzzy c-means (KFCM)| acoustic emission (AE) signals| low speed machine| roller element bearing

CLC Number:

TH 165

ZHANG Ke (张珂), LIN Tianran (林天然), JIN Xia (金霞) . Low Speed Bearing Fault Diagnosis Based on EMD-CIIT Histogram Entropy and KFCM Clustering[J]. Journal of Shanghai Jiao Tong University (Science), 2019, 24(5): 616-621.

References 20

[6]	PANDYA D H, UPADHYAY S H, HARSHA S P. Faultdiagnosis of rolling element bearing with intrinsic modefunction of acoustic emission data using APF-KNN [J].Expert Systems with Applications, 2013, 40(10): 4137-4145.
[7]	FENG Z P, LIANG M, CHU F L. Recent advances intime–frequency analysis methods for machinery faultdiagnosis: A review with application examples [J]. MechanicalSystems and Signal Processing, 2013, 38(1):165-205.
[1]	MBA D, RAO R B K N. Development of acoustic emissiontechnology for condition monitoring and diagnosisof rotating machines: bearings, pumps, gearboxes,engines, and rotating structures [J]. Shock and VibrationDigest, 2006, 38(1): 3-16.
[2]	SMITH J D. Vibration monitoring of bearings at lowspeeds [J]. Tribology International, 1982, 15(3): 139-144.
[8]	FENG Z P, LIANG M, ZHANG Y, et al. Faultdiagnosis for wind turbine planetary gearboxes via demodulationanalysis based on ensemble empirical modedecomposition and energy separation [J]. RenewableEnergy, 2012, 47: 112-126.
[9]	YU K, LIN T R, TAN J W. A bearing fault diagnosistechnique based on singular values of EEMD spatialcondition matrix and Gath-Geva clustering [J]. AppliedAcoustics, 2017, 121: 33-45.
[3]	YOSHIOKA T, FUJIWARA T. Application of acousticemission o detection of rolling bearing failure [J].ASME: Production Engineering Division Publication,1984, 14: 55-76.
[10]	LIN T R, KIM E, TAN A C C. A practical signalprocessing approach for condition monitoring of lowspeed machinery using Peak-Hold-Down-Sample algorithm[J]. Mechanical Systems and Signal Processing,2013, 36(2): 256-270.
[4]	MIETTINEN J, PATANIITTY P. Acoustic emissionin monitoring extremely slowly rotating rolling bearing[M]//MCINTYRE J, SLEEMAN D. Proceedingsof COMADEM’99. Oxford, UK: Coxmoor Publishing,1999: 289-297.
[11]	TIAN P F, ZHANG L, CAO X J, et al. The applicationof EMD-CIIT lidar signal denoising method in aerosoldetection [J]. Procedia Engineering, 2015, 102: 1233-1237.
[5]	VAN HECKE B, YOON J, HE D. Low speed bearingfault diagnosis using acoustic emission sensors [J].Applied Acoustics, 2016, 105: 35-44.
[12]	AO H, CHENG J S, LI K L, et al. A roller bearing faultdiagnosis method based on LCD energy entropy andACROA-SVM [J]. Shock and Vibration, 2014, 2014:825825.
[6]	PANDYA D H, UPADHYAY S H, HARSHA S P. Faultdiagnosis of rolling element bearing with intrinsic modefunction of acoustic emission data using APF-KNN [J].Expert Systems with Applications, 2013, 40(10): 4137-4145.
[13]	AI Y T, GUAN J Y, FEI C W, et al. Fusion informationentropy method of rolling bearing fault diagnosisbased on n-dimensional characteristic parameter distance[J]. Mechanical Systems and Signal Processing,2017, 88: 123-136.
[14]	ZHAO S F, LIANG L, XU G H, et al. Quantitativediagnosis of a spall-like fault of a rolling element bearingby empirical mode decomposition and the approximateentropy method [J]. Mechanical Systems and SignalProcessing, 2013, 40(1): 154-177.
[7]	FENG Z P, LIANG M, CHU F L. Recent advances intime–frequency analysis methods for machinery faultdiagnosis: A review with application examples [J]. MechanicalSystems and Signal Processing, 2013, 38(1):165-205.
[15]	ZHANG L, ZHANG L, HU J F, et al. Bearing faultdiagnosis using a novel classifier ensemble based onlifting wavelet packet transforms and sample entropy[J]. Shock and Vibration, 2016, 2016: 4805383.
[8]	FENG Z P, LIANG M, ZHANG Y, et al. Faultdiagnosis for wind turbine planetary gearboxes via demodulationanalysis based on ensemble empirical modedecomposition and energy separation [J]. RenewableEnergy, 2012, 47: 112-126.
[16]	ZACHARY J, IYENGAR S S, BARHEN J. Contentbased image retrieval and information theory: Ageneral approach [J]. Journal of the American Societyfor Information Science and Technology, 2001, 52(10):840-852.
[9]	YU K, LIN T R, TAN J W. A bearing fault diagnosistechnique based on singular values of EEMD spatialcondition matrix and Gath-Geva clustering [J]. AppliedAcoustics, 2017, 121: 33-45.
[10]	LIN T R, KIM E, TAN A C C. A practical signalprocessing approach for condition monitoring of lowspeed machinery using Peak-Hold-Down-Sample algorithm[J]. Mechanical Systems and Signal Processing,2013, 36(2): 256-270.
[17]	TAO X M, XU J, FU Q, et al. Kernel fuzzy C-meansalgorithm based on distribution density and its applicationin fault diagnosis [J]. Journal of Vibration andShock, 2009, 28(8): 61-64 (in Chinese).
[11]	TIAN P F, ZHANG L, CAO X J, et al. The applicationof EMD-CIIT lidar signal denoising method in aerosoldetection [J]. Procedia Engineering, 2015, 102: 1233-1237.
[18]	LIN T R, YU K, TAN J W. Condition monitoringand fault diagnosis of roller element bearing[EB/OL]. (2017-05-31) [2018-07-20]. https://www.intechopen.com.
[12]	AO H, CHENG J S, LI K L, et al. A roller bearing faultdiagnosis method based on LCD energy entropy andACROA-SVM [J]. Shock and Vibration, 2014, 2014:825825.
[19]	KOPSINIS Y, MCLAUGHLIN S. Development ofEMD-based denoising methods inspired by waveletthresholding [J]. IEEE Transactions on Signal Processing,2009, 57(4): 1351-1362.
[13]	AI Y T, GUAN J Y, FEI C W, et al. Fusion informationentropy method of rolling bearing fault diagnosisbased on n-dimensional characteristic parameter distance[J]. Mechanical Systems and Signal Processing,2017, 88: 123-136.
[20]	FLANDRIN P, RILLING G, GONCALVES P. Empiricalmode decomposition as a filter bank [J]. IEEESignal Processing Letters, 2004, 11(2): 112-114.
[14]	ZHAO S F, LIANG L, XU G H, et al. Quantitativediagnosis of a spall-like fault of a rolling element bearingby empirical mode decomposition and the approximateentropy method [J]. Mechanical Systems and SignalProcessing, 2013, 40(1): 154-177.
[15]	ZHANG L, ZHANG L, HU J F, et al. Bearing faultdiagnosis using a novel classifier ensemble based onlifting wavelet packet transforms and sample entropy[J]. Shock and Vibration, 2016, 2016: 4805383.
[16]	ZACHARY J, IYENGAR S S, BARHEN J. Contentbased image retrieval and information theory: Ageneral approach [J]. Journal of the American Societyfor Information Science and Technology, 2001, 52(10):840-852.
[17]	TAO X M, XU J, FU Q, et al. Kernel fuzzy C-meansalgorithm based on distribution density and its applicationin fault diagnosis [J]. Journal of Vibration andShock, 2009, 28(8): 61-64 (in Chinese).
[18]	LIN T R, YU K, TAN J W. Condition monitoringand fault diagnosis of roller element bearing[EB/OL]. (2017-05-31) [2018-07-20]. https://www.intechopen.com.
[19]	KOPSINIS Y, MCLAUGHLIN S. Development ofEMD-based denoising methods inspired by waveletthresholding [J]. IEEE Transactions on Signal Processing,2009, 57(4): 1351-1362.
[20]	FLANDRIN P, RILLING G, GONCALVES P. Empiricalmode decomposition as a filter bank [J]. IEEESignal Processing Letters, 2004, 11(2): 112-114.