Attention mechanism based multi-scale feature extraction of bearing fault diagnosis

doi:10.23919/JSEE.2023.000129

Journal of Systems Engineering and Electronics ›› 2023, Vol. 34 ›› Issue (5): 1359-1367.doi: 10.23919/JSEE.2023.000129

• Reliability • Previous Articles Next Articles

Attention mechanism based multi-scale feature extraction of bearing fault diagnosis

Xue LEI¹(), Ningyun LU¹^,²^,*(), Chuang CHEN¹^,³(), Tianzhen HU¹(), Bin JIANG¹^,²()

¹ College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
² State Key Laboratory of Mechanics and Control of Mechanical Structures, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
³ College of Electrical Engineering and Control Science, Nanjing Tech University, Nanjing 211816, China

Received:2021-08-16 Online:2023-10-18 Published:2023-10-30
Contact: Ningyun LU E-mail:leixue@nuaa.edu.cn;luningyun@nuaa.edu.cn;chenchuang@nuaa.edu.cn;hutianzhen@nuaa.edu.cn;binjiang@nuaa.edu.cn
About author:
LEI Xue was born in 1991. She received her M.S. degree in control theory and control engineering from Lanzhou University of Technology, Lanzhou, China, in 2018. She is currently pursuing her Ph.D. degree in control theory and control engineering with the College of Automation Engineering, Nanjing University of Aeronautics and Astronautics. Her current research interests include data-driven fault diagnosis, fault prognosis, and rotating machinery fault and health management. E-mail: leixue@nuaa.edu.cn

LU Ningyun was born in 1978. She received her Ph.D. degree from Northeastern University, Shenyang, China, in 2004. From 2002 to 2005, she worked as a research associate and post-doctoral fellow in Hong Kong University of Science and Technology. She is currently a full professor at the College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China. Her research interests include data-driven fault prognosis and diagnosis and their applications to various industrial processes. E-mail: luningyun@nuaa.edu.cn

CHEN Chuang was born in 1992. He received his M.Eng. degree in control engineering from Nanjing Tech University, Nanjing, China, in 2018, and Ph.D. degree in control theory and control engineering from Nanjing University of Aeronautics and Astronautics, Nanjing, China, in 2022. From March 2021 to April 2022, he became a China Scholarship Council (CSC)-funded joint Ph.D. student at York University, Canada. He is currently a lecturer in College of Electrical Engineering and Control Science at Nanjing Tech University, Nanjing, China. His current research interests include data-driven fault prognosis and health management. E-mail: chenchuang@nuaa.edu.cn

HU Tianzhen was born in 1996. She received her M.S. degree from Nanjing University of Aeronautics and Astronautics, Nanjing, China, in 2020. She is currently pursuing her Ph.D. degree in control science and engineering at Nanjing University of Aeronautics and Astronautics. Her research interests include data-driven fault diagnosis and fault propagation analysis. E-mail: hutianzhen@nuaa.edu.cn

JIANG Bin was born in 1966. He received his Ph.D. degree in automatic control from Northeastern University, Shenyang, China, in 1995. He had ever been a post-doctoral fellow, a research fellow, and a visiting professor in Singapore, France, USA and Canada, respectively. He is currently a chair professor of Cheung Kong Scholar Program in Ministry of Education, and the vice president of Nanjing University of Aeronautics and Astronautics, Nanjing, China. His current research interests include fault diagnosis and fault tolerant control and their applications in aircrafts, satellites and high-speed trains. E-mail: binjiang@nuaa.edu.cn
Supported by:
This work was supported by the National Natural Science Foundation of China (62020106003; 61873122; 62303217), Aero Engine Corporation of China Industry-university-research Cooperation Project (HFZL2020CXY011), and the Research Fund of State Key Laboratory of Mechanics and Control of Mechanical Structures (Nanjing University of Aeronautics and Astronautics) (MCMS-I-0121G03)

Abstract

Abstract:

Effective bearing fault diagnosis is vital for the safe and reliable operation of rotating machinery. In practical applications, bearings often work at various rotational speeds as well as load conditions. Yet, the bearing fault diagnosis under multiple conditions is a new subject, which needs to be further explored. Therefore, a multi-scale deep belief network (DBN) method integrated with attention mechanism is proposed for the purpose of extracting the multi-scale core features from vibration signals, containing four primary steps: preprocessing of multi-scale data, feature extraction, feature fusion, and fault classification. The key novelties include multi-scale feature extraction using multi-scale DBN algorithm, and feature fusion using attention mechanism. The benchmark dataset from University of Ottawa is applied to validate the effectiveness as well as advantages of this method. Furthermore, the aforementioned method is compared with four classical fault diagnosis methods reported in the literature, and the comparison results show that our proposed method has higher diagnostic accuracy and better robustness.

Key words: bearing fault diagnosis, multiple conditions, attention mechanism, multi-scale data, deep belief network (DBN)

Xue LEI, Ningyun LU, Chuang CHEN, Tianzhen HU, Bin JIANG. Attention mechanism based multi-scale feature extraction of bearing fault diagnosis[J]. Journal of Systems Engineering and Electronics, 2023, 34(5): 1359-1367.

Figures/Tables 18

Fig 1

Fig 2

Fig 3

Fig 4

Fig 5

Fig 6

Table 1

Fig 7

Fig 8

Table 2

Fig 9

Fig 10

Fig 11

Table 3

Fig 12

Fig 13

Fig 14

Fig 15

References 31

1	NANDI S, TOLIYAT H A, LI X D Condition monitoring and fault diagnosis of electrical motors—a review. IEEE Trans. on Energy Conversion, 2005, 20 (4): 719- 729. doi: 10.1109/TEC.2005.847955
2	MOSHREFZADEH A Condition monitoring and intelligent diagnosis of rolling element bearings under constant/variable load and speed conditions. Mechanical Systems and Signal Processing, 2021, 149, 107153. doi: 10.1016/j.ymssp.2020.107153
3	WANG B, LEI Y G, LI N, et al Multi-scale convolutional attention network for predicting remaining useful life of machinery. IEEE Trans. on Industrial Electronics, 2020, 68 (8): 7496- 7504.
4	WANG C S, LUN Y, CHEN Y H, et al. A data-driven aero-engine degradation prognostic strategy. IEEE Trans. on Cybernetics, 2021, 51(3): 1531−1541.
5	LEI Y G, YANG B, JIANG X W, et al Applications of machine learning to machine fault diagnosis: a review and roadmap. Mechanical Systems and Signal Processing, 2020, 138, 106587. doi: 10.1016/j.ymssp.2019.106587
6	LI X, JIA X D, ZHANG W, et al Intelligent cross-machine fault diagnosis approach with deep auto-encoder and domain adaptation. Neurocomputing, 2020, 383 (28): 235- 247.
7	CHEN Y H, PENG G L, ZHU Z Y, et al A novel deep learning method based on attention mechanism for bearing remaining useful life prediction. Applied Soft Computing, 2019, 86, 105919.
8	PATHIRAGE C S N, LI J, LI L, et al Application of deep autoencoder model for structural condition monitoring. Journal of Systems Engineering and Electronics, 2018, 29 (4): 873- 880. doi: 10.21629/JSEE.2018.04.22
9	CHEN H T, JIANG B, LU N Y Probability-relevant incipient fault detection and diagnosis methodology with applications to electric drive systems. IEEE Trans. on Control Systems Technology, 2018, 27 (6): 2766- 2773.
10	LI Y, ZIO E, LU N Y, et al Joint distribution-based test selection for fault detection and isolation under multiple faults condition. IEEE Trans. on Instrumentation and Measurement, 2020, 70, 1- 13.
11	CHEN C, LU N Y, JIANG B, et al Condition-based maintenance optimization for continuously monitored degrading systems under imperfect maintenance actions. Journal of Systems Engineering and Electronics, 2020, 31 (4): 841- 851. doi: 10.23919/JSEE.2020.000057
12	JAVED K, GOURIVEAU R, ZERHOUNI N, et al Enabling health monitoring approach based on vibration data for accurate prognostics. IEEE Trans. on Industrial Electronics, 2014, 62 (1): 647- 656.
13	NIU G X, WANG X, GOLDA M, et al An optimized adaptive PReLU-DBN for rolling element bearing fault diagnosis. Neurocomputing, 2021, 445, 26- 34. doi: 10.1016/j.neucom.2021.02.078
14	LI X Q, JIANG H K, WANG R X, et al Rolling bearing fault diagnosis using optimal ensemble deep transfer network. Knowledge-Based Systems, 2020, 213, 106695.
15	NASIRI A, TAHERI-GARAVAND A, OMID M, et al Intelligent fault diagnosis of cooling radiator based on deep learning analysis of infrared thermal images. Applied Thermal Engineering, 2019, 163, 114410. doi: 10.1016/j.applthermaleng.2019.114410
16	AZAMFAR M, LI X, LEE J Intelligent ball screw fault diagnosis using a deep domain adaptation methodology. Mechanism and Machine Theory, 2020, 151, 103932. doi: 10.1016/j.mechmachtheory.2020.103932
17	YANG Z B, ZHANG J P, ZHAO Z B, et al Interpreting network knowledge with attention mechanism for bearing fault diagnosis. Applied Soft Computing, 2020, 97, 106829. doi: 10.1016/j.asoc.2020.106829
18	JIAO J Y, ZHAO M, LIN J, et al A comprehensive review on convolutional neural network in machine fault diagnosis. Neurocomputing, 2020, 417, 36- 63. doi: 10.1016/j.neucom.2020.07.088
19	LI C, ZHANG S H, QIN Y, et al A systematic review of deep transfer learning for machinery fault diagnosis. Neurocomputing, 2020, 407, 121- 135. doi: 10.1016/j.neucom.2020.04.045
20	WANG Y R, SUN G D, JIN Q Imbalanced sample fault diagnosis of rotating machinery using conditional variational auto-encoder generative adversarial network. Applied Soft Computing, 2020, 92, 106333. doi: 10.1016/j.asoc.2020.106333
21	IRFAN M, SAAD N, IBRAHIM R, et al An intelligent fault diagnosis of induction motors in an arbitrary noisy environment. Journal of Nondestructive Evaluation, 2016, 35 (1): 1- 13. doi: 10.1007/s10921-015-0318-4
22	HINTONG E, SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks. Science, 2006, 313(5786): 504−507.
23	ZHONG T, QU J F, FANG X Y, et al The intermittent fault diagnosis of analog circuits based on EEMD-DBN. Neurocomputing, 2021, 436, 74- 91. doi: 10.1016/j.neucom.2021.01.001
24	LIU K, WU J K, LIU H B, et al Reliability analysis of thermal error model based on DBN and Monte Carlo method. Mechanical Systems and Signal Processing, 2021, 146, 107020. doi: 10.1016/j.ymssp.2020.107020
25	PAN Y B, HONG R G, CHEN J, et al A hybrid DBN-SOM-PF-based prognostic approach of remaining useful life for wind turbine gearbox. Renewable Energy, 2020, 152, 138- 154. doi: 10.1016/j.renene.2020.01.042
26	HU Q, SI X S, ZHANG Q H, et al A rotating machinery fault diagnosis method based on multi-scale dimensionless indicators and random forests. Mechanical Systems and Signal Processing, 2020, 139, 106609.1- 106609.22.
27	XU Z F, LI C, YANG Y Fault diagnosis of rolling bearings using an improved multi-scale convolutional neural network with feature attention mechanism. ISA Transactions, 2021, 110, 379- 393. doi: 10.1016/j.isatra.2020.10.054
28	JIANG J R, LEE J E, ZENG Y M. Time series multiple channel convolutional neural network with attention-based long short-term memory for predicting bearing remaining useful life. Sensors, 2019, 20(1): 166.
29	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
30	JIE H, LI S, GANG S, et al Squeeze-and-excitation networks. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2020, 42 (8): 2011- 2023. doi: 10.1109/TPAMI.2019.2913372
31	HUANG H, BADDOUR N, LIANG M Bearing fault diagnosis under unknown time-varying rotational speed conditions via multiple time-frequency curve extraction. Journal of Sound & Vibration, 2018, 414, 43- 60.

Speed varying conditions	Bearing health condition	Label	Training dataset	Testing dataset
IS	Healthy	1	10 500	4 500
	Inner race fault	2	10 500	4 500
	Outer race fault	3	10 500	4 500
DS	Healthy	4	10 500	4 500
	Inner race fault	5	10 500	4 500
	Outer race fault	6	10 500	4 500
IDS	Healthy	7	10 500	4 500
	Inner race fault	8	10 500	4 500
	Outer race fault	9	10 500	4 500
DIS	Healthy	10	10 500	4 500
	Inner race fault	11	10 500	4 500
	Outer race fault	12	10 500	4 500

Number of scales	Average accuracy/%	Average standard deviation	Average diagnostic time/s
1	96.26	2.041	311.594
2	96.48	0.749	538.051
3	98.73	0.153	767.237
4	98.48	0.166	1006.977
5	98.55	0.166	1226.475

Method	Speed varying condition
	Increasing speed		Decreasing speed		Increasing then decreasing speed		Combo time-varying rotational speed conditions
	Average accuracy/%	Standard deviation	Average accuracy/%	Standard deviation	Average accuracy/%	Standard deviation	Average accuracy/%	Standard deviation
Multi-scale DBN	98.73	0.153	99.041	0.178	99.886	0.083	96.593	0.515
Single-scale DBN	96.48	0.749	93.131	0.190	92.739	0.693	89.177	0.527
SVM	92.77	0.190	90.986	3.673	90.082	2.118	85.368	2.861
BPNN	84.79	2.341	89.499	2.413	87.717	3.231	85.202	3.588
PNN	86.13	1.390	91.686	1.817	91.517	1.611	81.900	2.366

Attention mechanism based multi-scale feature extraction of bearing fault diagnosis

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

Share this article

Figures/Tables 18

References 31

Related Articles 1

Recommended Articles

Metrics

Comments