Rolling bearing fault diagnostics based on improved data augmentation and ConvNet

doi:10.23919/JSEE.2023.000109

Journal of Systems Engineering and Electronics ›› 2023, Vol. 34 ›› Issue (4): 1074-1084.doi: 10.23919/JSEE.2023.000109

• RELIABILITY • Previous Articles

Rolling bearing fault diagnostics based on improved data augmentation and ConvNet

Delanyo Kwame Bensah KULEVOME¹^,²(), Hong WANG¹^,²^,*(), Xuegang WANG¹()

¹ School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China
² Yangtze Delta Region Institute (Huzhou), University of Electronic Science and Technology of China, Huzhou 313001, China

Received:2022-05-12 Online:2023-08-18 Published:2023-08-28
Contact: Hong WANG E-mail:kdelanyo@ieee.org;hongw@uestc.edu.cn;xgwang@uestc.edu.cn
About author:
KULEVOME Delanyo Kwame Bensah was born in 1983. He received his M.E. degree in electronic science and engineering in 2019 from the University of Electronic Science and Technology of China, Chengdu, China, where he is currently pursuing his Ph.D. degree in information and communication engineering. His research interests include prognostics and health management of systems, fault diagnostics, signal processing, and deep learning. E-mail: kdelanyo@ieee.org

WANG Hong was born in 1974. He received his B.S., M.S. and Ph.D. degrees from Northwestern Polytechnical University, Chongqing University, and University of Electronic Science and Technology of China (UESTC), respectively. He is a faculty member with UESTC since 2003. From 2007 to 2009 he was engaged in doctoral research with the Second Research Institute of Civil Aviation Administration. From 2009 to 2010, he was a research scholar with Polytechnic Institute of New York University and research assistant with New Jersey Institute of Technology, USA. His research interests include radar signal processing, avionics, aeronautical telecommunication, and surveillance technologies in air traffic control. E-mail: hongw@uestc.edu.cn

WANG Xuegang was born in 1962. He received his Ph.D. degree from Xidian University in 1992. He is now a professor and Ph.D. supervisor with University of Electronic Science and Technology of China. His research interests include radar signal processing, and millimeter wave radar. E-mail: xgwang@uestc.edu.cn
Supported by:
This work was supported by the National Natural Science Foundation of China (42027805), and the National Aeronautical Fund (ASFC-20172080005)

Abstract

Abstract:

Convolutional neural networks (CNNs) are well suited to bearing fault classification due to their ability to learn discriminative spectro-temporal patterns. However, gathering sufficient cases of faulty conditions in real-world engineering scenarios to train an intelligent diagnosis system is challenging. This paper proposes a fault diagnosis method combining several augmentation schemes to alleviate the problem of limited fault data. We begin by identifying relevant parameters that influence the construction of a spectrogram. We leverage the uncertainty principle in processing time-frequency domain signals, making it impossible to simultaneously achieve good time and frequency resolutions. A key determinant of this phenomenon is the window function’s choice and length used in implementing the short-time Fourier transform. The Gaussian, Kaiser, and rectangular windows are selected in the experimentation due to their diverse characteristics. The overlap parameter ’s size also influences the outcome and resolution of the spectrogram. A 50% overlap is used in the original data transformation, and ±25% is used in implementing an effective augmentation policy to which two-stage regular CNN can be applied to achieve improved performance. The best model reaches an accuracy of 99.98% and a cross-domain accuracy of 92.54%. When combined with data augmentation, the proposed model yields cutting-edge results.

Key words: bearing failure, short-time Fourier transform, prognostics and health management, data augmentation, fault diagnosis

Delanyo Kwame Bensah KULEVOME, Hong WANG, Xuegang WANG. Rolling bearing fault diagnostics based on improved data augmentation and ConvNet[J]. Journal of Systems Engineering and Electronics, 2023, 34(4): 1074-1084.

Figures/Tables 14

Fig 1

Fig 2

Fig 3

Fig 4

Fig 5

Table 1

Fig 6

Table 2

Experimental results of different combinations of real and augmented datasets % "

Data set	0 hp				1 hp				2 hp				3 hp
Data set	Acc	Pre	Rec	f1	Acc	Pre	Rec	f1	Acc	Pre	Rec	f1	Acc	Pre	Rec	f1
0	99.85	99.86	99.85	99.85	99.87	99.88	99.87	99.87	99.74	99.75	99.75	99.75	99.62	99.66	99.62	99.62
1	99.98	99.98	99.98	99.98	99.94	99.94	99.94	99.94	99.87	99.88	99.88	99.88	99.81	99.81	99.81	99.81
2	99.92	99.93	99.93	99.93	99.87	99.87	99.87	99.87	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94
3	99.79	99.79	99.79	99.79	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98
4	99.77	99.78	99.78	99.78	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98
5	99.77	99.78	99.78	99.78	99.62	99.65	99.62	99.62	99.87	99.88	99.88	99.88	99.98	99.98	99.98	99.98
6	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98
7	99.92	99.93	99.93	99.93	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98
8	99.92	99.93	99.93	99.93	99.81	99.81	99.81	99.81	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98
9	99.85	99.86	99.85	99.85	99.87	99.88	99.87	99.87	99.94	99.94	99.94	99.94	99.87	99.88	99.87	99.87
10	99.70	99.71	99.70	99.70	99.81	99.81	99.81	99.81	99.68	99.70	99.69	99.69	99.94	99.94	99.94	99.94
11	99.92	99.93	99.93	99.93	99.87	99.88	99.87	99.87	99.98	99.98	99.98	99.98	99.87	99.88	99.88	99.88
12	99.85	99.86	99.85	99.85	99.81	99.81	99.81	99.81	99.68	99.70	99.70	99.70	99.98	99.98	99.98	99.98
13	99.85	99.86	99.85	99.85	99.98	99.98	99.98	99.98	99.87	99.88	99.87	99.87	99.87	99.88	99.88	99.88
14	99.93	99.93	99.93	99.93	99.98	99.98	99.98	99.98	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98
15	99.92	99.93	99.93	99.93	99.94	99.94	99.94	99.94	99.74	99.78	99.78	99.78	99.98	99.98	99.98	99.98
16	99.85	99.86	99.85	99.85	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98	99.94	99.94	99.94	99.94
17	99.92	99.93	99.93	99.93	99.87	99.88	99.87	99.87	99.81	99.81	99.81	99.81	99.98	99.98	99.98	99.98
18	99.85	99.86	99.85	99.85	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98
19	99.85	99.86	99.85	99.85	99.81	99.82	99.81	99.81	99.87	99.88	99.88	99.88	99.98	99.98	99.98	99.98
20	99.92	99.93	99.93	99.93	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94	99.94
21	99.92	99.93	99.93	99.93	99.94	99.94	99.94	99.94	99.98	99.98	99.98	99.98	99.98	99.98	99.98	99.98

Table 2

Fig 7

Fig 8

Fig 9

Fig 10

Table 3

Table 4

References 32

1	GOUGAM F, RAHMOUNE C, BENAZZOUZ D, et al. Bearing faults classification under various operation modes using time domain features, singular value decomposition, and fuzzy logic system. Advances in Mechanical Engineering, 2020, 12(10): 1687814020967874.
2	ROHANI BASTAMI A, BASHARI A Rolling element bearing diagnosis using spectral kurtosis based on optimized impulse response wavelet. Journal of Vibration and Control, 2020, 26 (3/4): 175- 185. doi: 10.1177/1077546319877702
3	ZHANG Y, XING K S, BAI R X, et al An enhanced convolutional neural network for bearing fault diagnosis based on time-frequency image. Measurement, 2020, 157, 107667. doi: 10.1016/j.measurement.2020.107667
4	WANG Z Z, CHEN Y, CAI Z, et al Methods for predicting the remaining useful life of equipment in consideration of the random failure threshold. Journal of Systems Engineering and Electronics, 2020, 31 (2): 415- 431. doi: 10.23919/JSEE.2020.000018
5	KULEVOME D K B, WANG H, WANG X G A bidirectional LSTM-based prognostication of electrolytic capacitor. Progress in Electromagnetics Research C, 2021, 109, 139- 152. doi: 10.2528/PIERC20120201
6	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, 2012: 1097−1105.
7	AGBLEY B L Y, LI J, HAQ A, et al. Wavelet-based cough signal decomposition for multimodal classification. Proc. of the 17th International Computer Conference Wavelet Active Media Technology and Information Processing, 2020: 5–9.
8	AGBLEY B L Y, LI J, HAQ A U, et al. Multimodal melanoma detection with federated learning. Proc of the 18th International Computer Conference Wavelet Active Media Technology and Information Processing, 2021: 238–244.
9	ZHANG S, ZHANG S B, WANG B N, et al Deep learning algorithms for bearing fault diagnostics—a comprehensive review. IEEE Access, 2020, 8, 29857- 29881. doi: 10.1109/ACCESS.2020.2972859
10	KULEVOME D K B, WANG H, WANG X G Deep neural network based classification of rolling element bearings and health degradation through comprehensive vibration signal analysis. Journal of Systems Engineering and Electronics, 2022, 33 (1): 233- 246. doi: 10.23919/JSEE.2022.000023
11	CHEN X H, ZHANG B K, GAO D Bearing fault diagnosis base on multi-scale CNN and LSTM model. Journal of Intelligent Manufacturing, 2021, 32 (4): 971- 987. doi: 10.1007/s10845-020-01600-2
12	KUMAR A, ZHOU Y, GANDHI C P, et al Bearing defect size assessment using wavelet transform based deep convolutional neural network (DCNN). Alexandria Engineering Journal, 2020, 59 (2): 999- 1012. doi: 10.1016/j.aej.2020.03.034
13	YANG B, LEI Y G, JIA F, et al An intelligent fault diagnosis approach based on transfer learning from laboratory bearings to locomotive bearings. Mechanical Systems and Signal Processing, 2019, 122, 692- 706. doi: 10.1016/j.ymssp.2018.12.051
14	LI X, ZHANG W, DING Q, et al Intelligent rotating machinery fault diagnosis based on deep learning using data augmentation. Journal of Intelligent Manufacturing, 2020, 31 (2): 433- 452. doi: 10.1007/s10845-018-1456-1
15	LIU S W, JIANG H K, WU Z H, et al Data synthesis using deep feature enhanced generative adversarial networks for rolling bearing imbalanced fault diagnosis. Mechanical Systems and Signal Processing, 2022, 163, 108139. doi: 10.1016/j.ymssp.2021.108139
16	HU T H, TANG T, LIN R L, et al A simple data augmentation algorithm and a self-adaptive convolutional architecture for few-shot fault diagnosis under different working conditions. Measurement, 2020, 156, 107539. doi: 10.1016/j.measurement.2020.107539
17	GAO X, DENG F, YUE X H Data augmentation in fault diagnosis based on the Wasserstein generative adversarial network with gradient penalty. Neurocomputing, 2020, 396, 487- 494. doi: 10.1016/j.neucom.2018.10.109
18	AHMED H, NANDI A K. Condition monitoring with vibration signals: compressive sampling and learning algorithms for rotating machines. New Jersey: John Wiley & Sons, 2020.
19	IWANA B K, UCHIDA S An empirical survey of data augmentation for time series classification with neural networks. PLoS One, 2021, 16 (7): e0254841. doi: 10.1371/journal.pone.0254841
20	JIN K H, MCCANN M T, FROUSTEY E, et al Deep convolutional neural network for inverse problems in imaging. IEEE Trans. on Image Processing, 2017, 26 (9): 4509- 4522. doi: 10.1109/TIP.2017.2713099
21	NYKAZA E, BUNKLEY S, BLEVINS M G. Objectively choosing spectrogram parameters to classify environmental noises. INTER-NOISE and NOISE-CON Congress and Conference Proceedings, 2006, 253 (1): 7336–7343.
22	GAN J Z, LIU T, LI L, et al Non-negative matrix factorization: a survey. The Computer Journal, 2021, 64 (7): 1080- 1092. doi: 10.1093/comjnl/bxab103
23	BERRY M W, BROWNE M, LANGVILLE A N, et al Algorithms and applications for approximate nonnegative matrix factorization. Computational Statistics & Data Analysis, 2007, 52 (1): 155- 173.
24	COHEN L. Time-frequency analysis. New Jersey: Prentice Hall, 1995.
25	ALLEN R L, MILLS D. Signal analysis: time, frequency, scale, and structure. New Jersey: John Wiley & Sons, 2004.
26	SHAO H D, JIANG H K, LIN Y, et al A novel method for intelligent fault diagnosis of rolling bearings using ensemble deep auto-encoders. Mechanical Systems and Signal Processing, 2018, 102, 278- 297. doi: 10.1016/j.ymssp.2017.09.026
27	TIAN Y L, LIU X Y A deep adaptive learning method for rolling bearing fault diagnosis using immunity. Tsinghua Science and Technology, 2019, 24 (6): 750- 762. doi: 10.26599/TST.2018.9010144
28	UDMALE S S, SINGH S K Application of spectral kurtosis and improved extreme learning machine for bearing fault classification. IEEE Trans. on Instrumentation and Measurement, 2019, 68 (11): 4222- 4233. doi: 10.1109/TIM.2018.2890329
29	LI J M, YAO X F, WANG X D, et al Multiscale local features learning based on BP neural network for rolling bearing intelligent fault diagnosis. Measurement, 2020, 153, 107419. doi: 10.1016/j.measurement.2019.107419
30	GAN M, WANG C, ZHU C A Construction of hierarchical diagnosis network based on deep learning and its application in the fault pattern recognition of rolling element bearings. Mechanical Systems and Signal Processing, 2016, 72, 92- 104.
31	XU Z F, LI C, YANG Y Fault diagnosis of rolling bearing of wind turbines based on the variational mode decomposition and deep convolutional neural networks. Applied Soft Computing, 2020, 95, 106515. doi: 10.1016/j.asoc.2020.106515
32	LI X, ZHANG W, DING Q Understanding and improving deep learning-based rolling bearing fault diagnosis with attention mechanism. Signal Processing, 2019, 161, 136- 154. doi: 10.1016/j.sigpro.2019.03.019

Bearing operating condition	Diameter of faults/mm	Size of data								Class label
		Original		Overlap		Window		NMF
		${L_0}$	${L_{1,2,3}}$	${L_0}$	${L_{1,2,3}}$	${L_0}$	${L_{1,2,3}}$	${L_0}$	${L_{1,2,3}}$
Normal	0	480	960	960	1920	1440	2880	480	960	1
Ball	0.18	240	240	480	480	720	720	240	240	2
Ball	0.36	240	240	480	480	720	720	240	240	3
Ball	0.54	240	240	480	480	720	720	240	240	4
Inner race	0.18	240	240	480	480	720	720	240	240	5
Inner race	0.36	240	240	480	480	720	720	240	240	6
Inner race	0.54	240	240	480	480	720	720	240	240	7
Outer race	0.18	240	240	480	480	720	720	240	240	8
Outer race	0.36	240	240	480	480	720	720	240	240	9
Outer race	0.54	240	240	480	480	720	720	240	240	10

Method	Description	Number of classes	Training samples	Average test accuracy/%
Method 1 [26]	Ensemble deep auto-encoders	12	2400	97.18
Method 2 [27]	DCNN with frequency domain (FD) features	3	4500	99.38
Method 3 [28]	ELM with spectral Kurtosis	5	−	98.84
Method 4 [29]	Multiscale local feature learning with support vector machine	10	150	99.31
Method 5 [30]	Hierarchical diagnosis network with deep belief network	10	500	99.03
Our method	CNN with augmentation	10	5280 (A + G)	99.98

Test	Train
Test	0 hp	1 hp	2 hp	3 hp
0 hp	−	96.06	90.61	82.27
1 hp	95.83	−	96.54	87.18
2 hp	88.78	99.30	−	98.53
3 hp	83.21	92.69	99.49	−
Average accuracy	89.27	96.02	95.54	89.34

[1]	Xiaoyu WAN, Mingwei SHEN, Di WU, Daiyin ZHU. Wind turbine clutter mitigation using morphological component analysis with group sparsity [J]. Journal of Systems Engineering and Electronics, 2023, 34(3): 714-722.
[2]	Zhengyu YE, Bin JIANG, Yuehua CHENG, Ziquan YU, Yang YANG. Distributed fault diagnosis observer for multi-agent system against actuator and sensor faults [J]. Journal of Systems Engineering and Electronics, 2023, 34(3): 766-774.
[3]	Kwame Bensah KULEVOME Delanyo, Hong WANG, Xuegang WANG. Deep neural network based classification of rolling element bearings and health degradation through comprehensive vibration signal analysis [J]. Journal of Systems Engineering and Electronics, 2022, 33(1): 233-246.
[4]	Yuwei CUI, Aijun LI, Xianfeng MENG. A fault-tolerant control method for distributed flight control system facing wing damage [J]. Journal of Systems Engineering and Electronics, 2021, 32(5): 1041-1052.
[5]	Xiaogang QI, Bingchun WANG, Lifang LIU. Fault diagnosis based on dial-test data in datacenter networks [J]. Journal of Systems Engineering and Electronics, 2019, 30(5): 1035-1043.
[6]	Jiufu Liu, Wenliang Liu, Jianyong Zhou, Yan Sun, and Zhisheng Wang. Improved design of online fault diagnoser for partially observed Petri nets with generalized mutual exclusion constraints [J]. Systems Engineering and Electronics, 2017, 28(5): 971-978.
[7]	Bo Zhou, Kun Qian, Xudong Ma, and Xianzhong Dai. Ellipsoidal bounding set-membership identification approach for robust fault diagnosis with application to mobile robots [J]. Systems Engineering and Electronics, 2017, 28(5): 986-995.
[8]	Hui Sun, Jianguo Yan, Yaohong Qu, and Jie Ren. Sensor fault-tolerant observer applied in UAV anti-skid braking control under control input constraint [J]. Systems Engineering and Electronics, 2017, 28(1): 126-.
[9]	Junjie Huang, Zhen Jiang2, and Junwei Zhao. Component fault diagnosis for nonlinear systems [J]. Journal of Systems Engineering and Electronics, 2016, 27(6): 1283-1290.
[10]	Jing Qiu, Xiaodong Tan, Guanjun Liu, and Kehong L¨u. Test selection and optimization for PHM based on failure evolution mechanism model [J]. Journal of Systems Engineering and Electronics, 2013, 24(5): 780-792.
[11]	Rongling Lang, Zheping Xu, and Fei Gao. Data-driven fault diagnosis method for analog circuits based on robust competitive agglomeration [J]. Journal of Systems Engineering and Electronics, 2013, 24(4): 706-712.
[12]	Yuehua Cheng, Qian Hou, and Bin Jiang. Design and simulation of fault diagnosis based on NUIO/LMI for satellite attitude control systems [J]. Journal of Systems Engineering and Electronics, 2012, 23(4): 581-587.
[13]	Yun Li and Zhiming Cai. Gravity-based heuristic for set covering problems and its application in fault diagnosis [J]. Journal of Systems Engineering and Electronics, 2012, 23(3): 391-398.
[14]	Jiuping Xu and Lei Xu. Health management based on fusion prognostics for avionics systems [J]. Journal of Systems Engineering and Electronics, 2011, 22(3): 428-436.
[15]	Songyin Cao, Lei Guo, and Xinyu Wen. Robust fault diagnosis with disturbance rejection and attenuation for systems with multiple disturbances [J]. Journal of Systems Engineering and Electronics, 2011, 22(1): 135-140.

Rolling bearing fault diagnostics based on improved data augmentation and ConvNet

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

Share this article

Figures/Tables 14

References 32

Related Articles 15

Recommended Articles

Metrics

Comments