基于多源域深度迁移学习的机械故障诊断

杨胜康1,孔宪光1,王奇斌1,程涵1,李中权1,2

振动与冲击 ›› 2022, Vol. 41 ›› Issue (9) : 32-40.

PDF(2674 KB)
PDF(2674 KB)
振动与冲击 ›› 2022, Vol. 41 ›› Issue (9) : 32-40.
论文

基于多源域深度迁移学习的机械故障诊断

  • 杨胜康1,孔宪光1,王奇斌1,程涵1,李中权1,2
作者信息 +

Mechanical fault diagnosis based on multi-source domain deep transfer learning

  • YANG Shengkang1, KONG Xianguang1, WANG Qibin1, CHENG Han1, LI Zhongquan1,2
Author information +
文章历史 +

摘要

针对不同工况下的机械故障诊断问题,迁移学习方法相比于深度学习具有明显的成效,单源域迁移故障诊断仍会出现负迁移和模型泛化能力差的问题。因此,本文提出一种基于多源域深度迁移学习的机械故障诊断方法。首先,进行锚适配器的构建,获得多源域—目标域适配器数据对。其次,建立基于深度域适应的迁移学习网络模型获得每个数据对的分类器与预测结果。最后,采用加权集成的方式进行分类器集成,用于最终的故障诊断识别。本文所提方法充分集成多源域故障特征信息,提取域不变特征,避免负迁移的问题,提高模型的泛化能力。通过一个滚动轴承数据来验证提出方法的性能,结果表明,多工况迁移故障诊断分类精度明显高于其中任意单一工况迁移,最高可提高8.78%,与其他方法相比,本文所提方法具有较好的精度和泛化能力。

Abstract

For mechanical fault diagnosis under different working conditions, the transfer learning method has obvious effect compared with deep learning method, but the single source domain transfer fault diagnosis still has negative transfer and poor model generalization ability. Therefore, this paper proposes a machine fault diagnosis method based on multi-source domain deep transfer learning. Firstly, the anchor adapter is constructed to obtain the multi-source domain-target domain adapter data pairs. Secondly, the classifier and prediction results of each data pair are obtained based on the deep domain adaption transfer learning network model. Finally, the weighted integration method is adopted to ensemble classifier for the final fault diagnosis and identification. The proposed method fully ensembles multi-source domain fault feature information, extracts domain invariant features, avoids negative transfer problems, and improves the generalization ability of the model. A rolling bearing dataset is used to verify the performance of the proposed method. The results show that the fault diagnosis classification accuracy of multi-condition migration is significantly higher than that of any single condition migration, and the maximum improvement is 8.78%. Compared with other methods, the proposed method has better accuracy and generalization ability.

关键词

故障诊断 / 多源域迁移学习 / 锚适配器集成 / 深度神经网络

Key words

Fault diagnosis / Multi-source domain transfer learning / Ensemble of anchor adapters / Deep neural network

引用本文

导出引用
杨胜康1,孔宪光1,王奇斌1,程涵1,李中权1,2. 基于多源域深度迁移学习的机械故障诊断[J]. 振动与冲击, 2022, 41(9): 32-40
YANG Shengkang1, KONG Xianguang1, WANG Qibin1, CHENG Han1, LI Zhongquan1,2. Mechanical fault diagnosis based on multi-source domain deep transfer learning[J]. Journal of Vibration and Shock, 2022, 41(9): 32-40

参考文献

[1] 雷亚国, 贾峰, 孔德同, 等. 大数据下机械智能故障诊断的机遇与挑战[J].机械工程学报, 2018, 54(05): 94-104.
 LEI Yaguo, JIA Feng, KONG Detong, et al. Opportunities and challenges of mechanical intelligent fault diagnosis under big data[J]. Journal of Mechanical Engineering, 2018, 54(05): 94-104.
[2] 赵学智, 叶邦彦, 陈统坚. 多分辨奇异值分解理论及其在信号处理和故障诊断中的应用[J]. 机械工程学报, 2010, 46(20): 64-75.
 ZHAO Xuezhi, YE Bangyan, CHEN Tongjian. Multi-resolution singular value decomposition and its application in signal processing and fault diagnosis[J]. Journal of Mechanical Engineering, 2010, 46(20): 64-75.
[3] 孟宗, 李姗姗. 基于小波改进阈值去噪和HHT的滚动轴承故障诊断[J]. 振动与冲击, 2013, 32(14): 204-208.
 MENG Zong, LI Shanshan. Fault diagnosis of rolling bearing based on wavelet improved threshold denoising and HHT[J]. Journal of Vibration and Shock. 2013, 32(14): 204-208.
[4] 王国彪, 何正嘉, 陈雪峰, 等. 机械故障诊断基础研究“何去何从”[J]. 机械工程学报, 2013, 49(1): 63-72.
 WANG Guobiao, HE Zhengjia, CHEN Xuefeng, et al. Basic research on machinery fault diagnosis—what is the prescription[J]. Journal of Mechanical Engineering, 2013, 49(1): 63-72.
[5] Xu Yansun, WEAVER John B, HEALY Dennis M, et al. Wavelet transform domain filters: a spatially selective noise filtration technique[J]. IEEE Trans Image Process, 1994, 3(6): 747-758.
[6] 张琛, 赵荣珍, 邓林峰. 基于EEMD奇异值熵的滚动轴承故障诊断方法[J]. 振动、测试与诊断, 2019, 39(02): 353-358+446-447.
 ZHANG Chen, ZHAO Rongzhen, DENG Linfeng. Rolling bearing fault diagnosis method based on EEMD singular value entropy[J]. Journal of Vibration, Measurement & Diagnosis, 2019, 39(02): 353-358+446-447.
[7] 王梦阳, 王华庆, 董方,等. 基于EVMD-LNMF的复合故障信号分离方法[J]. 振动与冲击, 2019, 38(16): 145-152.
 WANG Mengyang, WANG Huaqing, DONG Fang, et al. Compound fault signal separation method based on EVMD-LNMF[J]. Journal of Vibration and Shock, 2019, 38(16): 146-152.
[8] 张少敏,毛冬,王保义.大数据处理技术在风电机组齿轮箱故障诊断与预警中的应用[J].电力系统自动化, 2016, 40(14): 129-134.
 ZHANG Shaomin, MAO Dong, WANG Baoyi. Application of big data processing technology in fault diagnosis and early warning of wind turbine gearbox[J]. Automation of electric power systems, 2016, 40(14): 129-134.
[9] NI Jianjun, ZHANG Chuanbiao, YANG Simon X. An adaptive approach based on KPCA and SVM for real-time fault diagnosis of HVCBs[J]. IEEE Transactions on Power Delivery, 2011, 26(3): 1960-1971.
[10] KORBICZ Jozef. Artificial neural networks in fault diagnosis of dynamical systems[C]//2010 IEEE Region 8 International Conference on Computational Technologies in Electrical and Electronics Engineering (SIBIRCON). IEEE, 2010: 449-449.
[11] LI Chuan, SANCHEZ R V, ZURITA Grover, et al. Gearbox fault diagnosis based on deep random forest fusion of acoustic and vibratory signals[J]. Mechanical Systems And Signal Processing, 2016, 76-77: 283-293.
[12] SHAO Haidong, JIANG Hongkai, Zhang Xun, et al. Rolling bearing fault diagnosis using an optimization deep belief network[J]. Measurement Science and Technology, 2015, 26(11): 115002.
[13] ZHOU Funa, YANG Shuai, WEN Chenglin, et al. Improved DAE and application in fault diagnosis[C]. Chinese Control And Decision Conference, 2018.
[14] 李涛, 段礼祥, 张东宁, 等. 自适应卷积神经网络在旋转机械故障诊断中的应用[J]. 振动与冲击, 2020, 39(16): 275-282+288.
 LI Tao, DUAN Lixiang, ZHANG Dongning, et al. Application of adaptive convolutional neural network in rotating machinery fault diagnosis[J]. Journal of Vibration and Shock, 2020, 39(16): 275-282+288.
[15] PAN Sinno Jialin, YANG Qiang. A survey on transfer learning[J]. IEEE Transactions on knowledge and data engineering, 2009, 22(10): 1345-1359.
[16] WEN Long, GAO Liang, LI Xinyu. A new deep transfer learning based on sparse auto-encoder for fault diagnosis[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2017, 49(1): 136-144.
[17] YANG Bin, LEI Yaguo, JIA Feng, et al. An intelligent fault diagnosis approach based on transfer learning from laboratory bearings to locomotive bearings[J]. Mechanical Systems and Signal Processing, 2019, 122: 692-706.
[18] ZHANG Ran, TAO Hongyang, WU Lifeng, et al. Transfer learning with neural networks for bearing fault diagnosis in changing working conditions[J]. IEEE Access, 2017, 5: 14347-14357.
[19] XU Kun, LI Shunming, WANG Jinrui, et al. A novel convolutional transfer feature discrimination network for unbalanced fault diagnosis under variable rotational speeds[J]. Measurement Science and Technology, 2019, 30(10): 105107.
[20] LI Xiang, ZHANG Wei, DING Qian, et al. Diagnosing rotating machines with weakly supervised data using deep transfer learning, IEEE Transactions on Industrial Informatics, 2020 16(3) 1688-1697.
[21] Zhang Y, Ren Z, Zhou S. An intelligent fault diagnosis for rolling bearing based on adversarial semi-supervised method[J]. IEEE Access, 2020, 8: 149868-149877.
[22] Zhang W, Li X, Ma H, et al. Federated learning for machinery fault diagnosis with dynamic validation and self-supervision[J]. Knowledge-Based Systems, 2021, 213.
[23] Li J, Huang R, He G, et al. A two-stage transfer adversarial network for intelligent fault diagnosis of rotating machinery with multiple new faults[J]. IEEE/ASME Transactions on Mechatronics, 2020: 1-1.
[24] Huang R, Wang Z, Li J, et al. A Transferable Capsule Network for Decoupling Compound Fault of Machinery[C]. 2020 IEEE International Instrumentation and Measurement Technology Conference (I2MTC), 2020: 1-6.
[25] Zhuang F, Luo P, Pan S J, et al. Ensemble of Anchor Adapters for Transfer Learning[C]. Proceedings of the 25th ACM International on Conference on Information and Knowledge Management, 2016: 2335–2340.
[26] Case Western Reserve University Bearing Data Center, Available: http://csegroups.case.edu/bearingdatacenter/home.
[27] Xu K, Li S, Wang J, et al. A novel convolutional transfer feature discrimination network for unbalanced fault diagnosis under variable rotational speeds[J]. Measurement Science and Technology, 2019, 30(10): 105107.
[28] Ganin Y, Ustinova E, Ajakan H, et al. Domain-Adversarial Training of Neural Networks[J].Journal of Machine Learning Research, 2017, 17(1):2096-2030.

PDF(2674 KB)

Accesses

Citation

Detail

段落导航
相关文章

/