伪标签驱动局部子空间对齐的跨域故障诊断方法

摘要
图/表
参考文献(24)
相关文章 (15)

全文: PDF (1816 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要由于不同工况下的同类故障特征间存在分布差异，导致了源域中训练的智能诊断模型在应用至目标域时出现性能退化。针对此问题，提出一种伪标签驱动局部子空间对齐，以减少不同域分布差异从而实现跨域故障诊断的方法。首先通过迁移源域中的预训练模型并联合余弦相似度的计算，在模型不同位置为目标域无标签样本计算样本伪标签的概率分布；然后引入局部最大平均差异 (Local maximum mean difference, LMMD)来减少源域和目标域相同子空间的特征分布偏差，实现对齐源域和目标域相关子空间；并且采用宽核卷积神经网络 (Convolutional neural network, CNN) 提取深层次跨域不变特征，最终实现较高故障诊断准确率的跨域智能故障诊断。实验结果表明，在两组验证数据集上所提方法均实现了最高的故障诊断准确率，证明了该方法的优越性，具有较好的实际应用价值。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	张猛1
	王波1
	2
	徐浩1
	杨文龙1
	汪超1

关键词 ：故障诊断, 跨域, 伪标签, 局部子空间, 卷积神经网络

Abstract：The distribution differences among similar fault features under different operating conditions lead to the performance degradation of the intelligent diagnostic model trained in the source domain when applied to the target domain.. To break the predicament, a pseudo-label driven local subspace alignment method for cross-domain fault diagnosis is proposed. Firstly, by migrating the pre-trained model and combining it with the cosine similarity calculation, the pseudo-label probability distributions of the unlabeled samples in the target domain are calculated at different locations of the model. Then, local maximum mean difference (LMMD) is introduced to reduce the deviation of feature distribution in the same subspace of source and target domains, in order to align the relevant subspaces of source and target domains. Furthermore, the convolutional neural network (CNN) with wide convolution kernels is used to extract deeper cross-domain invariant features, which realizes cross-domain intelligent fault diagnosis with high fault diagnosis accuracy. The proposed method achieves the highest fault diagnosis accuracy on both validation data sets, which entirely proves the superiority of the method and has preferable application value.

Key words： fault diagnosis cross-domain pseudo-label local subspace convolutional neural network

收稿日期: 2022-08-02 出版日期: 2023-10-28

引用本文:

张猛1，王波1,2，徐浩1，杨文龙1，汪超1. 伪标签驱动局部子空间对齐的跨域故障诊断方法[J]. 振动与冲击, 2023, 42(20): 105-113.
ZHANG Meng1,WANG Bo1,2,XU Hao1,YANG Wenlong1,WANG Chao1. A cross-domain fault diagnosis method based on pseudo-label driving local subspace alignment. JOURNAL OF VIBRATION AND SHOCK, 2023, 42(20): 105-113.

链接本文:

http://jvs.sjtu.edu.cn/CN/ 或 http://jvs.sjtu.edu.cn/CN/Y2023/V42/I20/105

[1] 康守强，王玉静，杨广学，等. 基于经验模态分解和超球多类支持向量机的滚动轴承故障诊断方法[J]. 中国电机工程学报，2011, 31(14): 96-102.
Kang Shou-qiang, Wang Yu-jing, Yang Guang-xue, et al. Rolling Bearing Fault Diagnosis Method Using Empirical Mode Decomposition and Hypersphere Multiclass Support Vector Machine[J]. Proceedings of the CSEE, 2011, 31(14): 96-102.
[2] Zhao R, Yan R Q, Chen Z H, et al. Deep learning and its applications to machine health monitoring[J]. Mechanical Systems and Signal Processing. 2019, 115: 213-237.
[3] Yang B, Xu S C, Lei Y G, et al. Multi-source transfer learning network to complement knowledge for intelligent diagnosis of machines with unseen faults[J]. Mechanical Systems and Signal Processing, 2022, 162: 108095.
[4] 王波，刘树林，蒋超，等. 基于量子遗传算法优化RVM的滚动轴承智能故障诊断[J]. 振动与冲击, 2015, 34(17): 207-212.
WANG Bo, LIU Shu-lin, JIANG Chao, et al. Rolling bearing intelligent fault diagnosis based on RVM optimized by Quantum genetic algorithm. JOURNAL OF VIBRATION AND SHOCK, 2015, 34(17): 207-212.
[5] Guo L, Lei Y G, Xing S B, et al. Deep Convolutional Transfer Learning Network: A New Method for Intelligent Fault Diagnosis of Machines With Unlabeled Data[J]. IEEE Transactions on Industrial Electronics, 2019, 66(9): 7316-7325.
[6] Li X, Zhang W, Ding Q, et al. Intelligent rotating machinery fault diagnosis based on deep learning using data augmentation[J]. Journal of Intelligent Manufacturing. 2020, 31(2): 433-452.
[7] Zhao Z B, Zhang Q Y, Yu X L, et al. Applications of Unsupervised Deep Transfer Learning to Intelligent Fault Diagnosis: A Survey and Comparative Study[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-28.
[8] Lu N N, Xiao H H, Sun Y J, et al. A new method for intelligent fault diagnosis of machines based on unsupervised domain adaptation[J]. Neurocomputing, 2021, 427: 96-109.
[9] 陈祝云，钟琪，黄如意，等. 基于增强迁移卷积神经网络的机械智能故障诊断[J]. 机械工程学报, 2021, 57(21): 96-105.
CHEN Zhu-yun, ZHONG Qi, HUANG Ru-yi, et al. Intelligent Fault Diagnosis for Machinery Based on Enhanced Transfer Convolutional Neural Network[J]. Journal of Mechanical Engineering, 2021, 57(21): 96-105.
[10] He J, Ouyang M, Chen Z W, et al. A Deep Transfer Learning Fault Diagnosis Method Based on WGAN and Minimum Singular Value for Non-Homologous Bearing[J]. IEEE Transactions on Instrumentation and Measurement, 2022, 71: 1-9.
[11] Liu X Y, Liu S L, Xiang J W, et al. An ensemble and shared selective adversarial network for partial domain fault diagnosis of machinery[J]. Engineering Applications of Artificial Intelligence. 2022, 113: 104906.
[12] Lu W N, Liang B, Cheng Y, et al. Deep Model Based Domain Adaptation for Fault Diagnosis[J]. IEEE Transactions on Industrial Electronics, 2017, 64(3): 2296-2305.
[13] Wang B, Wang B Q, Ning Y. A novel transfer learning fault diagnosis method for rolling bearing based on feature correlation matching[J]. Measurement Science and Technology. 2022, 33(12): 125006.
[14] 胡若晖，张敏，许文鑫. 基于DCGAN和DANN网络的滚动轴承跨域故障诊断[J]. 振动与冲击，2022, 41(06): 21-29.
HU Ruo-hui, ZHANG Min, XU Wen-xin. Cross-domain fault diagnosis of rolling element bearings using DCGAN and DANN. JOURNAL OF VIBRATION AND SHOCK, 2022, 41(6): 21-29.
[15] Wang B, Ning Y, Zhang Y H. A novel fault diagnosis scheme for rolling bearing based on symbolic aggregate approximation and convolutional neural network with channel attention[J]. Measurement Science and Technology. 2022, 33(1): 15016.
[16] Zhu Y C, Zhuang F Z, Wang J D, et al. Deep Subdomain Adaptation Network for Image Classification[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(4): 1713-1722.
[17] Song X D, Cong Y Y, Song Y F, et al. A bearing fault diagnosis model based on CNN with wide convolution kernels[J]. Journal of Ambient Intelligence and Humanized Computing, 2021.
[18] Wang J R, Li S M, An Z H, et al. Batch-normalized deep neural networks for achieving fast intelligent fault diagnosis of machines[J]. Neurocomputing, 2019, 329: 53-65.
[19] Chen E, Chen V. In-sensor time-domain classifiers using pseudo sigmoid activation functions[J]. Integration, 2020, 73: 43-49.
[20] Xu G S, Wu H Z, Shi Y Q. Structural Design of Convolutional Neural Networks for Steganalysis[J]. IEEE Signal Processing Letters, 2016, 23(5): 708-712.
[21] Karmakar S, Mukherjee A. Provable training of a ReLU gate with an iterative non-gradient algorithm[J]. Neural Networks, 2022, 151: 264-275.
[22] Zhang B, Li W, Li X L, et al. Intelligent Fault Diagnosis Under Varying Working Conditions Based on Domain Adaptive Convolutional Neural Networks[J]. IEEE Access, 2018, 6: 66367-66384.
[23] Che C C, Wang H W, Ni X M, et al. Domain adaptive deep belief network for rolling bearing fault diagnosis[J]. Computers & Industrial Engineering, 2020, 143: 106427.
[24] Zhang Y C, Ren Z H, Zhou S H. A New Deep Convolutional Domain Adaptation Network for Bearing Fault Diagnosis under Different Working Conditions[J]. Shock and Vibration, 2020, 2020: 1-14.