Bearing Fault Diagnosis Method Based on Intensive Distribution Alignment with Open Set Difference

doi:10.3969/j.issn.1004-132X.2024.09.012

Abstract

Abstract: In view of the facts that the current transfer learning algorithm was applied to the field of bearing fault diagnosis, there might be unknown fault categories in the target domain, which led to the problems of low recognition accuracy, and an open set domain adaptive method was introduced. Aiming at the problems that the existing open set domain adaptive algorithm paid more attentions to cross-domain alignment and paid little attention to the distribution within the domain when performing distribution alignment, which led to the low recognition rate of unknown categories. A method was proposed based on intensive distribution alignment with open set difference(IDAOD), which integrated cross-domain divergence alignment and intra-domain distribution alienation method. Firstly, the open set nearest neighbor class verification method was used to obtain the pseudo-label of the target domain. Then, the overall divergence matrix of the source domain and the target domain was constructed, and the cross-domain divergence alignment was performed. Based on the distribution adaptation weighted conditional distribution, the spatial distribution of different categories was further alienated in the same domain. Finally, under the framework of structural risk minimization, the loss function constructed was introduced based on the open set difference theory, and the regularization term was introduced to obtain the optimal solution and the new target domain pseudo-label. The feasibility of the IDAOD algorithm was verified by experiments on Office-31 dataset. The fault diagnosis experiments were carried out on CWRU and JNU bearing data sets respectively, and the recognition rate of unknown fault category is higher than that of other comparative open set algorithms. It is verified that the method proposed herein may effectively improve the accuracy of the bearing data set when target sample contains unknown samples.

Key words: open set difference, distribution alignment, bearing fault diagnosis, divergence matrix

摘要： 针对当前迁移学习算法应用于轴承故障诊断领域时，目标域中可能存在未知故障类别进而导致识别准确率过低的问题，引入了开集域自适应方法。针对现有的开放集域自适应算法在进行分布对齐时，多注重跨域对齐，很少关注领域内部的分布情况，从而导致对未知类别识别率较低的问题，提出了一种采用集约化开放集差异分布对齐（IDAOD）策略的轴承故障诊断方法，集成了跨域散度对齐和域内分布异化方法。使用开放集最近邻类验证方法获取目标域伪标签，然后构造源域和目标域的总体散度矩阵，进行跨域的散度对齐。基于分布适配加权条件分布，进一步异化同领域内不同类别的空间分布。在结构风险最小化框架下，基于开放集差异理论构造的损失函数，引入正则化项，得到最优解和新的目标域伪标签。通过在Office-31数据集上进行试验，验证了IDAOD算法的可行性。分别在CWRU和JNU轴承数据集上进行故障诊断试验，所提方法对未知故障类别的识别率均高于其他对比开放集算法，验证了所提方法在轴承数据集上能有效提高目标样本含有未知样本时的准确率。

关键词: 开放集差异, 分布对齐, 轴承故障诊断, 散度矩阵

CLC Number:

TH17

LI Can1, 2, WANG Guangbin2, 1, ZHAO Shubiao1, 2, ZHONG Zhixian1, ZHANG Hui1, 2. Bearing Fault Diagnosis Method Based on Intensive Distribution Alignment with Open Set Difference[J]. China Mechanical Engineering, 2024, 35(09): 1622-1633.

李灿1, 2, 王广斌2, 1, 赵树标1, 2, 钟志贤1, 张慧1, 2. 采用集约化开放集差异分布对齐策略的轴承故障诊断方法[J]. 中国机械工程, 2024, 35(09): 1622-1633.

References

［1］黄南天, 杨学航, 蔡国伟, 等. 采用非平衡小样本数据的风机主轴承故障深度对抗诊断［J］. 中国电机工程学报, 2020, 40(2):563-574.
HUANG Nantian, YANG Xuehang, CAI Guowei, et al. A Deep Adversarial Diagnosis Method for Wind Turbine Main Bearing Fault with Imbalanced Small Sample Scenarios［J］. Proceedings of the CSEE, 2020, 40(2):563-574.
［2］XIAO Ni, ZHANG Lei. Dynamic Weighted Learning for Unsupervised Domain Adaptation［C］∥2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). IEEE, 2021:15237-15246.
［3］温江涛, 张鹏程, 孙洁娣, 等. 残差卷积自编码网络无监督迁移轴承故障诊断［J］. 中国机械工程, 2022, 33(14):1707-1716.
WEN Jiangtao, ZHANG Pengcheng, SUN Jiedi,et al. Unsupervised Transfer Learning with Residual Convolutional Autoencoder Networks for Bearing Fault Diagnosis［J］. China Mechanical Engineering, 2022, 33(14):1707-1716.
［4］LI Yuanyuan, SHENG Hanmin, CHENG Yuhua, et al. State-of-health Estimation of Lithium-ion Batteries Based on Semi-supervised Transfer Component Analysis［J］. Applied Energy, 2020, 277:115504.
［5］HAN Te, LIU Chao, YANG Wenguang, et al. Deep Transfer Network with Joint Distribution Adaptation:a New Intelligent Fault Diagnosis Framework for Industry Application［J］. ISA Transactions, 2020, 97:269-281.
［6］DANG Wei, XIANG Longhai, LIU Shan, et al. A Feature Matching Method Based on the Convolutional Neural Network［J］. Journal of Imaging Science and Technology, 2023, 67(3):030402.
［7］MENG Yu, XUAN Jianping, XU Long, et al. Dynamic Reweighted Domain Adaption for Cross-domain Bearing Fault Diagnosis［J］. Machines, 2022, 10(4):245.
［8］吴磊, 王家序, 张新, 等. 基于最大重加权峭度盲解卷积的风电故障诊断［J］. 中国机械工程, 2022, 33(19):2356-2363.
WU Lei, WANG Jiaxu, ZHANG Xin, et al.Blind Deconvolution Based on Reweighted-kurtosis Maximization for Wind Turbine Fault Diagnosis［J］. China Mechanical Engineering, 2022, 33(19):2356-2363.
［9］BEN-DAVID S, BLITZER J, CRAMMER K, et al. Analysis of Representations for Domain Adaptation［M］∥SCHLKOPF B， PLATT J， HOFMANN T. Advances in Neural Information Processing Systems 19.Cambridge:MIT Press, 2007:137-144.
［10］HU Qin, SI Xiaosheng, QIN Aisong, et al. Balanced Adaptation Regularization Based Transfer Learning for Unsupervised Cross-domain Fault Diagnosis［J］. IEEE Sensors Journal, 2022, 22(12):12139-12151.
［11］WANG Tianchi, YU Lu, WANG Wenyu, et al. Specific Emitter Identification Based on the Multi-discrepancy Deep Adaptation Network［J］. IET Radar, Sonar & Navigation, 2022, 16(12):2079-2088.
［12］SHEN Jian, QU Yanru, ZHANG Weinan, et al. Wasserstein Distance Guided Representation Learning for Domain Adaptation［C］∥Proceedings of the AAAI Conference on Artificial Intelligence. New Orleans, 2018:10.1609/aaai.v32i1.11784.
［13］董绍江, 朱朋, 朱孙科, 等. 基于仿真数据驱动和领域自适应的滚动轴承故障诊断方法［J］. 中国机械工程, 2023, 34(6):694-702.
DONG Shaojiang, ZHU Peng, ZHU Sunke, et al. Fault Diagnosis Method of Rolling Bearings Based on Simulation Data Drive and Domain Adaptation［J］. China Mechanical Engineering, 2023, 34(6):694-702.
［14］BUSTO P P, GALL J. Open Set Domain Adaptation［C］∥2017 IEEE International Conference on Computer Vision(ICCV).Venice, 2017:754-763.
［15］SAITO K, YAMAMOTO S, USHIKU Y, et al. Open Set Domain Adaptation by Backpropagation［C］∥European Conference on Computer Vision—ECCV 2018. Cham:Springer International Publishing, 2018:156-171.
［16］ZHONG Li, FANG Zhen, LIU Feng, et al. Bridging the Theoretical Bound and Deep Algorithms for Open Set Domain Adaptation［J］. IEEE Transactions on Neural Networks and Learning Systems, 2023, 34(8):3859-3873.
［17］LUO Yadan, WANG Zijian, CHEN Zhuoxiao, et al. Source-free Progressive Graph Learning for Open-set Domain Adaptation［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(9):11240-11255.
［18］WU Jun, HE Jingrui. Domain Adaptation with Dynamic Open-set Targets［C］∥Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining.Washington DC, 2022:2039-2049.
［19］FANG Zhen, LU Jie, LIU Feng, et al. Open Set Domain Adaptation:Theoretical Bound and Algorithm［J］. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(10):4309-4322.
［20］ZHAO Shubiao, WANG Guangbin, LI Xuejun, et al. Fault Diagnosis of Rolling Bearing Based on Cross-domain Divergence Alignment and Intra-domain Distribution Alienation［J］. Journal of Vibroengineering, 2023, 25(6):1124-1140.
［21］秦玉峰, 史贤俊. 基于MMD的故障可诊断性定量评价方法［J］. 控制与决策, 2023, 38(10):2925-2933.
QIN Yufeng, SHI Xianjun. Quantitative Evaluation Approach of Fault Diagnosability Based on Maximum Mean Discrepancy［J］. Control and Decision, 2023, 38(10):2925-2933.
［22］CHEN Renxiang, ZHU Jukun, HU Xiaolin, et al. Fault Diagnosis Method of Rolling Bearing Based on Multiple Classifier Ensemble of the Weighted and Balanced Distribution Adaptation under Limited Sample Imbalance［J］. ISA Transactions, 2021, 114:434-443.
［23］MENDES JNIOR P R, de SOUZA R M, de O WERNECK R, et al. Nearest Neighbors Distance Ratio Open-set Classifier［J］. Machine Learning, 2017, 106(3):359-386.
［24］WANG Guangbin, ZHAO Shubiao, ZHONG Zhixian, et al. Research on Shaft Current Damage Identification of Variable Condition Motor Bearings Based on Multiscale Feature Label Propagation and Manifold Metric Transfer［J］. Lubricants, 2023, 11(2):69.
［25］WANG Jindong, FENG Wenjie, CHEN Yiqiang, et al. Visual Domain Adaptation with Manifold Embedded Distribution Alignment［C］∥Proceedings of the 26th ACM International Conference on Multimedia.Seoul, 2018:402-410.
［26］ZHAO Huimin, ZHENG Jianjie, DENG Wu, et al.Semi-supervised Broad Learning System Based on Manifold Regularization and Broad Network［J］. IEEE Transactions on Circuits and Systems I:Regular Papers, 2020, 67(3):983-994.
［27］王军辉, 雷文平, 刘华杰, 等. 基于深度动态域适应的轴承故障诊断研究［J］. 振动与冲击, 2023, 42(14):245-250.
WANG Junhui, LEI Wenping, LIU Huajie, et al. Bearing Fault Diagnosis Based on Deep Dynamic Domain Adaptation［J］. Journal of Vibration and Shock, 2023, 42(14):245-250.

[1]	ZHAO Jing, YANG Shaopu, LI Qiang, LIU Yongqiang, . A New Transfer Learning Method with Residual Attention and Its Applications on Rolling Bearing Fault Diagnosis [J]. China Mechanical Engineering, 2023, 34(03): 332-343.
[2]	WEN Jiangtao, ZHANG Pengcheng, SUN Jiedi, LEI Ming. Unsupervised Transfer Learning with Residual Convolutional Autoencoder Networks for Bearing Fault Diagnosis [J]. China Mechanical Engineering, 2022, 33(14): 1707-1716.
[3]	HE Qiang, TANG Xianghong, LI Chuanjiang, LU Jianguang, CHEN Jiadui. Bearing Fault Diagnosis Method Based on Small Sample Data under Unbalanced Loads#br# [J]. China Mechanical Engineering, 2021, 32(10): 1164-1171,1180.
[4]	Zhang Shuqing1;Li Wei1;Zhang Liguo1，2;Hu Yongtao1;Qian Lei1;Jiang Wanlu1. Bearing Fault Diagnosis Based on multi-EMD, cApEn and GG Clustering Algorithm [J]. China Mechanical Engineering, 2016, 27(24): 3362-3367.