数据隐私与数据安全问题逐渐受到社会关注，各用户隐私的滚动轴承振动数据存在孤岛且不共享的问题，同时不同规格滚动轴承振动数据分布差异大、部分已知标签数据稀缺，使得诊断准确率不高。针对上述问题，提出一种基于联邦模型迁移的不同规格滚动轴承故障诊断框架。该方法对多个用户振动数据做短时傅里叶变换，构建时频图数据集；各用户训练本地模型并将模型参数上传至服务器，同时引入差值更新和参数稀疏化算法改进联邦学习中本地模型参数传递策略；服务器采用联邦平均算法聚合模型参数并更新本地模型，迭代后建立用于迁移学习的共享模型；提出逐层解冻策略保留共享模型部分参数并发送给每个用户，再利用本地数据微调共享模型，获得适用于每个用户的个性化模型。经实验验证，所提方法在数据孤岛和标签稀缺的前提下，可实现不同规格滚动轴承故障诊断，并具有较高的准确率和良好的泛化性能。
Abstract
The issues of data privacy and security have gained attention gradually, for each user, the private vibration data of rolling bearings existing the problems of isolated data island and the non-sharing of data among different users. Meanwhile, the distributions of vibration data for different specification rolling bearings have great difference, and some labeled data is scarce. So, above reasons lead to low diagnostic accuracy. For the above problems, a fault diagnosis frame of different specifications rolling bearings based on federated model transfer learning is proposed. The vibration data of multiple users are processed using short-time Fourier transform, and then time-frequency spectrum datasets can be constructed. Each user trains the local model and uploads model parameters to the server, the local model parameters transfer strategy of federated learning can be improved by introducing difference update and parameter thinning algorithm. By using the federated averaging algorithm, the server aggregates local model parameters and updates the local model, then after iterating, a shared model for transfer learning can be obtained. The server uses layer-by-layer unfreeze strategy to reserve part of the shared model parameters and send them to each user, and users fine-tune the shared model using local data, then getting a personalized model that works for each user. Through the experimental verification, the proposed method can realize fault diagnosis of rolling bearings for different specifications in the case of isolated data island and lack of labels, it also has high accuracy and good generalization performance.
关键词
联邦学习 /
迁移学习 /
滚动轴承 /
故障诊断 /
不同规格
{{custom_keyword}} /
Key words
federated learning /
transfer learning /
rolling bearing /
fault diagnosis /
different specifications
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 康守强，邹佳悦，王玉静，等. 基于无监督特征对齐的变负载下滚动轴承故障诊断方法[J]. 中国电机工程学报，2020, 40(01):274-281+393.
KANG Shou-qiang, ZOU Jia-yue, WANG Yu-jing, et al. Fault Diagnosis Method of a Rolling Bearing Under Varying Loads Based on Unsupervised Feature Alignment[J]. Proceedings of the CSEE, 2020, 40(8): 40(01): 274-281+393.
[2] 王玉静，吕海岩，康守强，等. 不同型号滚动轴承故障诊断方法[J]. 中国电机工程学报，2021, 41(01): 267-276+416.
WANG Yu-jing, LYU Hai-yan, KANG Shou-qiang, et al. Fault Diagnosis Method for Different Types of Rolling Bearings[J]. Proceedings of the CSEE, 2021, 41(01): 267-276+416.
[3] Zhang S, Zhang S, Wang B, et al. Deep learning algorithms for bearing fault diagnostics—A comprehensive review[J]. IEEE Access, 2020, 8: 29857-29881.
[4] 田科位，董绍江，姜保军，等. 基于改进深度残差网络的轴承故障诊断方法[J]. 振动与冲击，2021, 40(20): 247-254.
TIAN Ke-wei, DONG Shao-jiang, JIANG Bao-jun, et al. A bearing fault diagnosis method based on an improved depth residual network[J]. Journal of Vibration and Shock, 2021, 40(20): 247-254.
[5] 丁 雪，邓艾东，李 晶，等. 基于多尺度和注意力机制的滚动轴承故障诊断[J]. 东南大学学报（自然科学版），2021, 41(01): 267-276+416.
Ding Xue, Deng Ai-dong, Li Jing, et al. Fault diagnosis of helicopter rolling bearing based on improved SqueezeNet[J]. JOURNAl OF SOUTHEAST UNIVERSITY (Natural Science Edition), 2022, 52(01): 172-178.
[6] 余志锋，熊邦书，李新民，等. 基于改进的SqueezeNet直升机滚动轴承故障诊断[J]. 航空动力学报:1-10[2022-04-06].
YU Zhi-feng, XIONG Bang-shu, LI Xin-min, et al. Fault diagnosis of helicopter rolling bearing based on improved SqueezeNet[J]. Journal of Aerospace Power: 1-10[2022-04-06].
[7] Fu L, Zhang L, Tao J. An improved deep convolutional neural network with multiscale convolution kernels for fault diagnosis of rolling bearing[C].//IOP Conference Series:Materials Science and Engineering. IOP Publishing,2021,1043(5):052021.
[8] 张 龙，甄灿壮，熊国良，等. 基于深度时频特征的机车轴承故障诊断[J]. 交通运输工程学报, 2021,21(06): 247-258.
ZHANG Long, ZHEN Can-zhuang, XIONG Guo-liang, et al. Locomotive bearing fault diagnosis based on deep time-frequency features[J]． Journal of Traffic and Transportation Engineeering, 2021, 21(06): 247-258.
[9] 崔石玉，朱志宇. 基于参数迁移和一维卷积神经网络的海水泵故障诊断[J]. 振动与冲击，2021, 40(24): 180-189.
CUI Shi-yu, ZHU Zhi-yu. Seawater pump fault diagnosis based on parameter transfer and one-dimensional convolutional neural network[J]. Journal of Vibration and Shock, 2021, 40(24): 180-189.
[10] 张西宁，余 迪，刘书语. 基于迁移学习的小样本轴承故障诊断方法研究[J]. 西安交通大学学报，2021, 55(10): 30-37.
ZHANG Xi-ning, YU Di, LIU Shu-yu. Fault Diagnosis Method for Small Sample Bearing Based on Transfer Learning[J]. Journal of Xi’an Jiaotong University, 2021, 55(10): 30-37.
[11] 侯东晓，穆金涛，方 成，等. 基于GADF与引入迁移学习的ResNet34对变速轴承的故障诊断[J]. 东北大学学报（自然科学版），2022, 43(03): 383-389.
HOU Dong-xiao, MU Jin-tao, FANG Cheng, et al. Fault Diagnosis of Variable Speed Bearings Based on GADF and ResNet34 Introduced Transfer Learning[J]. Journal of Northeastern University(Natural Science), 2022, 43(03): 383-389.
[12] Jiang F, Shen X, Jiang F, et al. A bearing fault diagnosis method based on improved convolution neural network and transfer learning[C]//International Conference on Intelligent Equipment and Special Robots (ICIESR 2021).SPIE, 2021, 12127: 124-130.
[13] Fan H, Xue C, Zhang X, et al. Vibration images-driven fault diagnosis based on CNN and transfer learning of rolling bearing under strong noise[J]. Shock and Vibration, 2021, 2021.
[14]邵海东，肖一鸣，颜 深. 仿真数据驱动的改进无监督域适应轴承故障诊断[J/OL]. 机械工程学报: 1-10[2022-09-16].http://kns.cnki.net/kcms/detail/11.2187.TH.20220915.1441.002.html
Shao Hai-dong, Xiao Yi-ming, Yan Shen. Simulation Data-driven Enhanced Unsupervised Domain Adaptation for Bearing Fault Diagnosis[J/OL]. Journal of Mechanical Engineering: 1-10[2022-09-16].http://kns.cnki.net/kcms/detail/11.2187.TH.20220915.1441.002.html
[15] Yang Q, Liu Y, Cheng Y, et al. Federated learning[J]. Synthesis Lectures on Artificial Intelligence and Machine Learning, 2019, 13(3): 1-207.
[16] Geng D Q, He H W, Lan X C, et al. Bearing fault diagnosis based on improved federated learning algorithm[J]. Computing, 2021: 1-19.
[17] Li Z, Li Z, Li Y, et al. An Intelligent Diagnosis Method for Machine Fault Based on Federated Learning[J]. Applied Sciences, 2021, 11(24): 12117.
[18] Zhang Z, Xu X, Gong W, et al. Efficient federated convolutional neural network with information fusion for rolling bearing fault diagnosis[J]. Control Engineering Practice, 2021, 116: 104913.
[19] Zhou F, Zhang Z, Li S. Research on federated learning method for fault diagnosis in multiple working conditions[J]. Complex Engineering Systems, 2021, 1(2): 7.
[20] Zhang W, Li X, Ma H. Federated learning for machinery fault diagnosis with dynamic validation and self-supervision[J]. Knowledge-Base Systems, 2021, 213: 106679.
[21] Alzubaidi L, Zhang J, Humaidi A J, et al. Review of deep learning: Concepts, CNN architectures, challenges, applications, future directions[J]. Journal of big Data, 2021, 8(1): 1-74.
[22] KENNETH A Loparo. Bearing data center[M/OL]. Cleveland：Case Western Reserve University,2013.http://csegroups.case.edu/bearingdatacenter/home.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(2206 KB)
