A New Diagnosis Method with Few-shot Learning Based on a Class-rebalance Strategy for Scarce Faults in Industrial Processes

Xinyao Xu; De Xu; Fangbo Qin

doi:10.1007/s11633-022-1363-y

Article Contents

Xinyao Xu, De Xu, Fangbo Qin. A New Diagnosis Method with Few-shot Learning Based on a Class-rebalance Strategy for Scarce Faults in Industrial Processes. Machine Intelligence Research. https://doi.org/10.1007/s11633-022-1363-y

Citation:

Xinyao Xu, De Xu, Fangbo Qin. A New Diagnosis Method with Few-shot Learning Based on a Class-rebalance Strategy for Scarce Faults in Industrial Processes. Machine Intelligence Research. https://doi.org/10.1007/s11633-022-1363-y

Citation:

PDF( 1677 KB)

A New Diagnosis Method with Few-shot Learning Based on a Class-rebalance Strategy for Scarce Faults in Industrial Processes

doi: 10.1007/s11633-022-1363-y

Xinyao Xu^{1, 2

,},
De Xu^{1, 2

,},
Fangbo Qin^{1

,}

1.
Research Center of Precision Sensing and Control, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China
2.
School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 100049, China

More Information

Author Bio:
Xinyao Xu received the B. Sc. degree in automation from Tianjin University, China in 2018. He is currently a Ph. D. degree candidate in control science and engineering at Research Center of Precision Sensing and Control, Institute of Automation, Chinese Academy of Sciences (CASIA), and School of Artificial Intelligence, University of Chinese Academy of Sciences (UCAS), China. His research interests include deep learning, industrial fault detection, and fault diagnosis. E-mail: xuxinyao2018@ia.ac.cn ORCID iD: 0000-0002-6371-1948

De Xu received the B. Sc. and M. Sc. degrees from Shandong University of Technology, China in 1985 and 1990, respectively, and the Ph. D. degree from Zhejiang University, China in 2001, all in control science and engineering. He has been with CASIA since 2001. He is currently a professor with Research Center of Precision Sensing and Control, CASIA, China. He is also with School of Artificial Intelligence, UCAS, China. His research interests include robotics and automation such as visual measurement, visual control, intelligent control, visual positioning, microscopic vision, micro-assembly, and skill learning. E-mail: de.xu@ia.ac.cn (Corresponding author) ORCID iD: 0000-0002-7221-1654

Fangbo Qin received the B. Sc. degree in automation from Beijing Jiaotong University, China in 2013, the Ph. D. degree in control science and engineering from CASIA and UCAS, China in 2019. He is currently an associate professor with Research Center of Precision Sensing and Control, CASIA, China. His research interests include robot vision, robot manipulation, and deep learning. E-mail: qinfangbo2013@ia.ac.cn ORCID iD: 0000-0002-4085-0857
Received Date: 2022-04-22
Accepted Date: 2022-08-01
Publish Online: 2023-02-18

Abstract

Abstract

For industrial processes, new scarce faults are usually judged by experts. The lack of instances for these faults causes a severe data imbalance problem for a diagnosis model and leads to low performance. In this article, a new diagnosis method with few-shot learning based on a class-rebalance strategy is proposed to handle the problem. The proposed method is designed to transform instances of the different faults into a feature embedding space. In this way, the fault features can be transformed into separate feature clusters. The fault representations are calculated as the centers of feature clusters. The representations of new faults can also be effectively calculated with few support instances. Therefore, fault diagnosis can be achieved by estimating feature similarity between instances and faults. A cluster loss function is designed to enhance the feature clustering performance. Also, a class-rebalance strategy with data augmentation is designed to imitate potential faults with different reasons and degrees of severity to improve the model′s generalizability. It improves the diagnosis performance of the proposed method. Simulations of fault diagnosis with the proposed method were performed on the Tennessee-Eastman benchmark. The proposed method achieved average diagnosis accuracies ranging from 81.8% to 94.7% for the eight selected faults for the simulation settings of support instances ranging from 3 to 50. The simulation results verify the effectiveness of the proposed method.
- Data augmentation,
- feature clustering,
- class-rebalance strategy,
- few-shot learning,
- fault diagnosis

FullText(HTML)

References(35)

References

[1]	C. H. Hu, J. Y. Luo, X. Y. Kong, X. W. Feng. Novel fault subspace extraction methods for the reconstruction-based fault diagnosis. Journal of Process Control, vol. 105, pp. 129–140, 2021. DOI: 10.1016/j.jprocont.2021.07.008.
[2]	P. Zhou, R. Y. Zhang, J. Xie, J. P. Liu, H. Wang, T. Y. Chai. Data-driven monitoring and diagnosing of abnormal furnace conditions in blast furnace ironmaking: An integrated PCA-ICA method. IEEE Transactions on Industrial Electronics, vol. 68, no. 1, pp. 622–631, 2021. DOI: 10.1109/TIE.2020.2967708.
[3]	L. Wen, X. Y. Li, L. Gao, Y. Y. Zhang. A new convolutional neural network-based data-driven fault diagnosis method. IEEE Transactions on Industrial Electronics, vol. 65, no. 7, pp. 5990–5998, 2018. DOI: 10.1109/TIE.2017.2774777.
[4]	H. Liu, J. Z. Zhou, Y. Zheng, W. Jiang, Y. C. Zhang. Fault diagnosis of rolling bearings with recurrent neural network-based autoencoders. ISA Transactions, vol. 77, pp. 167–178, 2018. DOI: 10.1016/j.isatra.2018.04.005.
[5]	S. Yin, S. X. Ding, X. C. Xie, H. Luo. A review on basic data-driven approaches for industrial process monitoring. IEEE Transactions on Industrial Electronics, vol. 61, no. 11, pp. 6418–6428, 2014. DOI: 10.1109/TIE.2014.2301773.
[6]	N. Laouti, S. Othman, M. Alamir, N. Sheibat-Othman. Combination of model-based observer and support vector machines for fault detection of wind turbines. International Journal of Automation and Computing, vol. 11, no. 3, pp. 274–287, 2014. DOI: 10.1007/s11633-014-0790-9.
[7]	Y. Zhang, C. Bingham, M. Garlick, M. Gallimore. Applied fault detection and diagnosis for industrial gas turbine systems. International Journal of Automation and Computing, vol. 14, no. 4, pp. 463–473, 2017. DOI: 10.1007/s11633-016-0967-5.
[8]	H. Q. Wang, Y. L. Ke, G. G. Luo, G. Tang. Compressed sensing of roller bearing fault based on multiple down-sampling strategy. Measurement Science and Technology, vol. 27, no. 2, Article number 025009, 2016. DOI: 10.1088/0957-0233/27/2/025009.
[9]	N. V. Chawla, K. W. Bowyer, L. O. Hall, W. P. Kegelmeyer. SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, vol. 16, pp. 321–357, 2002. DOI: 10.1613/jair.953.
[10]	J. Mathew, C. K. Pang, M. Luo, W. H. Leong. Classification of imbalanced data by oversampling in kernel space of support vector machines. IEEE Transactions on Neural Networks and Learning Systems, vol. 29, no. 9, pp. 4065–4076, 2018. DOI: 10.1109/TNNLS.2017.2751612.
[11]	H. B. He, Y. Bai, E. A. Garcia, S. T. Li. ADASYN: Adaptive synthetic sampling approach for imbalanced learning. In Proceedings of IEEE International Joint Conference on Neural Networks, Hong Kong, China, pp. 1322-1328, 2008. DOI: 10.1109/IJCNN.2008.4633969.
[12]	X. X. Wu, Y. G. He, J. J. Duan. A deep parallel diagnostic method for transformer dissolved gas analysis. Applied Sciences, vol. 10, no. 4, Article number 1329, 2020. DOI: 10.3390/app10041329.
[13]	Q. W. Guo, Y. B. Li, Y. Song, D. C. Wang, W. Chen. Intelligent fault diagnosis method based on full 1-D convolutional generative adversarial network. IEEE Transactions on Industrial Informatics, vol. 16, no. 3, pp. 2044–2053, 2020. DOI: 10.1109/TII.2019.2934901.
[14]	Y. Zhuo, Z. Q. Ge. Auxiliary information-guided industrial data augmentation for any-shot fault learning and diagnosis. IEEE Transactions on Industrial Informatics, vol. 17, no. 11, pp. 7535–7545, 2021. DOI: 10.1109/TII.2021.3053106.
[15]	A. Odena, C. Olah, J. Shlens. Conditional image synthesis with auxiliary classifier GANs. In Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia, pp. 2642–2651, 2017.
[16]	Z. Y. Wu, W. F. Lin, Y. Ji. An integrated ensemble learning model for imbalanced fault diagnostics and prognostics. IEEE Access, vol. 6, pp. 8394–8402, 2018. DOI: 10.1109/ACCESS.2018.2807121.
[17]	L. J. Zhou, J. W. Dang, Z. H. Zhang. Fault classification for on-board equipment of high-speed railway based on attention capsule network. International Journal of Automation and Computing, vol. 18, no. 5, pp. 814–825, 2021. DOI: 10.1007/s11633-021-1291-2.
[18]	Z. X. Hu, P. Jiang. An imbalance modified deep neural network with dynamical incremental learning for chemical fault diagnosis. IEEE Transactions on Industrial Electronics, vol. 66, no. 1, pp. 540–550, 2019. DOI: 10.1109/TIE.2018.2798633.
[19]	W. K. Yu, C. H. Zhao. Broad convolutional neural network based industrial process fault diagnosis with incremental learning capability. IEEE Transactions on Industrial Electronics, vol. 67, no. 6, pp. 5081–5091, 2020. DOI: 10.1109/TIE.2019.2931255.
[20]	G. I. Parisi, R. Kemker, J. L. Part, C. Kanan, S. Wermter. Continual lifelong learning with neural networks: A review. Neural Networks, vol. 113, pp. 54–71, 2019. DOI: 10.1016/j.neunet.2019.01.012.
[21]	J. Deng, W. Dong, R. Socher, L. J. Li, K. Li, F. F. Li, ImageNet: A large-scale hierarchical image database. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Miami, USA, pp. 248–255, 2009. DOI: 10.1109/CVPR.2009.5206848.
[22]	J. Xu, P. F. Xu, Z. C. Wei, X. Ding, L. Shi. DC-NNMN: Across components fault diagnosis based on deep few-shot learning. Shock and Vibration, vol. 2020, Article number 3152174, 2020. DOI: 10.1155/2020/3152174.
[23]	C. J. Li, S. B. Li, A. S. Zhang, Q. He, Z. H. Liao, J. J. Hu. Meta-learning for few-shot bearing fault diagnosis under complex working conditions. Neurocomputing, vol. 439, pp. 197–211, 2021. DOI: 10.1016/j.neucom.2021.01.099.
[24]	A. S. Zhang, S. B. Li, Y. X. Cui, W. L. Yang, R. Z. Dong, J. J. Hu. Limited data rolling bearing fault diagnosis with few-shot learning. IEEE Access, vol. 7, pp. 110895–110904, 2019. DOI: 10.1109/ACCESS.2019.2934233.
[25]	N. Lu, H. Y. Hu, T. Yin, Y. G. Lei, S. H. Wang. Transfer relation network for fault diagnosis of rotating machinery with small data. IEEE Transactions on Cybernetics, to be published. DOI: 10.1109/TCYB.2021.3085476.
[26]	C. Finn, P. Abbeel, S. Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia, pp. 1126–1135, 2017.
[27]	D. Wang, M. Zhang, Y. C. Xu, W. N. Lu, J. Yang, T. Zhang. Metric-based meta-learning model for few-shot fault diagnosis under multiple limited data conditions. Mechanical Systems and Signal Processing, vol. 155, Article number 107510, 2021. DOI: 10.1016/j.ymssp.2020.107510.
[28]	J. Snell, K. Swersky, R. Zemel. Prototypical networks for few-shot learning. In Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, USA, pp. 4080−4090, 2017.
[29]	O. Vinyals, C. Blundell, T. Lillicrap, K. Kavukcuoglu, D. Wierstra. Matching networks for one shot learning. In Proceedings of the 30th International Conference on Neural Information Processing Systems, Barcelona, Spain, pp. 3637–3645, 2016.
[30]	G. Koch, R. Zemel, R. Salakhutdinov. Siamese neural networks for one-shot image recognition. In Proceedings of the 32nd International Conference on Machine Learning, Lille, France, 2015.
[31]	D. W. Li, Y. J. Tian. Survey and experimental study on metric learning methods. Neural Networks, vol. 105, pp. 447–462, 2018. DOI: 10.1016/j.neunet.2018.06.003.
[32]	F. Schroff, D. Kalenichenko, J. Philbin. FaceNet: A unified embedding for face recognition and clustering. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Boston, USA, pp. 815-823, 2015. DOI: 10.1109/CVPR.2015.7298682.
[33]	J. J. Downs, E. F. Vogel. A plant-wide industrial process control problem. Computers &Chemical Engineering, vol. 17, no. 3, pp. 245–255, 1993. DOI: 10.1016/0098-1354(93)80018-I.
[34]	A. Bathelt, N. L. Ricker, M. Jelali. Revision of the Tennessee Eastman process model. IFAC-PapersOnLine, vol. 48, no. 8, pp. 309–314, 2015. DOI: 10.1016/j.ifacol.2015.08.199.
[35]	L. Van der Maaten, G. Hinton. Visualizing data using t-SNE. Journal of Machine Learning Research, vol. 9, no. 86, pp. 2579–2605, 2008.

Relative Articles

Supplements(0)

Cited By

Proportional views

Proportional views

通讯作者: 陈斌, bchen63@163.com

1.
沈阳化工大学材料科学与工程学院沈阳 110142

Figures(9) / Tables(5)

Get Citation

PDF

XML

用微信扫码二维码

分享至好友和朋友圈

Article Metrics

Article views (15) PDF downloads(2)

A New Diagnosis Method with Few-shot Learning Based on a Class-rebalance Strategy for Scarce Faults in Industrial Processes

doi: 10.1007/s11633-022-1363-y

Abstract

References

Proportional views

Catalog

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Proportional views

Related

A New Diagnosis Method with Few-shot Learning Based on a Class-rebalance Strategy for Scarce Faults in Industrial Processes

doi: 10.1007/s11633-022-1363-y

Abstract

References

Proportional views

Catalog

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Proportional views

Related

Export File

Citation

Format

Content