Abstract:This paper proposes a bearing fault diagnosis approach based on one-dimensional convolution attention gated recurrent network (1DCNN-Attention-GRU) and transfer learning to improve the accuracy and generalization capability of the fault diagnosis model with small sample size. Firstly, a fault diagnosis network based on a one-dimensional convolution network (1DCNN), gated recurrent unit (GRU), and attention mechanism is constructed to decrease the dependence of traditional fault diagnosis methods on the artificial experience. Then, transfer learning is introduced, which trains the fault diagnosis model with massive source datasets and freezes the basic structure of the model, and then obtains the final model by fine-tuning its top-level structure using less target domain data. Finally, the Softmax function is employed for fault classification. The experimental results show the fault diagnosis accuracy of the proposed approach is higher than the result of 1DCNN-GRU, GRU, and support vector machine(SVM). In addition, when 3% of target domain data is used for fine-tuning, the fault diagnosis accuracy reaches 98% for various bearing working conditions, which indicates the proposed fault diagnosis method with transfer learning can classify various bearing working conditions with small sample size.
石静雯,侯立群. 基于一维卷积注意力门控循环网络和迁移学习的轴承故障诊断[J]. 振动与冲击, 2023, 42(3): 159-164.
SHI Jingwen, HOU Liqun. Bearing fault diagnosis based on 1D CNN attention gated recurrent network and transfer learning. JOURNAL OF VIBRATION AND SHOCK, 2023, 42(3): 159-164.
[1] M. M. M. Islam, J. Kim. Automated bearing fault diagnosis scheme using 2D representation of waveletpacket transform and deep convolutional neural network[J]. Computers in Industry, 106(2019): 142-153.
[2] 陈剑,蔡坤奇,陶善勇等.基于IITD模糊熵与随机森林的滚动轴承故障诊断方法[J].计量学报,2021,(06):774-779.
CHEN J,CAI K Q,TAO S Y,et al.Fault Diagnosisi Method of Rolling Bearing Based on IITD Fuzzy Entropy and Random Forest[J]. Acta Metrologica Sinica,2021,(06):774-779.
[3] 曹宁,江志农,高金吉.小样本下基于迁移学习的轴承状态识别方法[J].噪声与振动控制,2020,(05):89-94+132.
CAO N,JIANG Z N ,GAO J J.Rolling Bearing State Recognition Transfer Learning under Small Samples[J].Noise and Vibration Control,2020,(05): 89-94+132.
[4] 唐刚,朱立军,胡雄.基于优化VMD和深度置信网络的滚动轴承故障诊断[J].轴承,2020,(10):47-53.
TANG G,ZHU L J,HU X.Fault Diagnosis for Rolling Bearings Based on Optimized VMDand Deep Belief Network[J].Bearing2020,(10):47-53.
[5] 齐咏生,刘飞,李永亭等.基于MK-MOMEDA和Teager能量算子的风电机组滚动轴承复合故障诊断[J].太阳能学报,2021, (07):42.
QI Y S,LIU FLI Y T,et al.Compound Fault Diagnosis of Wind Turbing Rolling Bearing based on MK-MOMEDA and Teager Energy Operator[J].Acta Energiae Solaris Sinica,2021,(07):42.
[6] Zou F, Zhang H F, Sang S T,et al.Bearing fault diagnosis based on combined multi-scale weighted entropy morphological filtering and bi-LSTM[J].Measurement,2021,(02):176.
[7] Zhao J, Yang S P, Li Q.A new bearing fault diagnosis method based on signal-to-image mapping and convolutional neural network[J].Measurement,2021,(03):176.
[8] PAN S J, YANG Q.Asurvey on transfer learning[J].IEEE Transactions,2010,22(10):1345-1359.
[9] 李俊卿,刘静.结合卷积神经网络和迁移学习的电机轴承故障诊断方法[J].C,2021,(08):1-9/
[10] 苏靖涵,张潇.基于深度迁移自编码器的变工况下滚动轴承故障诊断方法[EB/OL].(2021-08-05). https://kns.cnki.net/kcms/detail/13.1212.TM.20210805.1136.002.html.
[11] Chen Z H,Cen J, Xiong J B.Rolling Bearing Fault Diagnosis Using Time-Frequency Analysis and Deep Transfer Convolutional Neural Network[J].IEEE ACCESS,2020,(08): 150248-150261.
[12] Shao J , Huang Z W, Zhu J N.Transfer Learning Method Based on Adversarial Domain Adaption for Bearing Fault Diagnosis[J]. IEEE ACCESS,2020,(06):51375289-51775323.
[13] 弗朗索瓦肖莱Python深度学习[M].张亮,译.北京:人民邮电出版社,2018:181-182.
[14] 任欢,王旭光.注意力机制综述[J].计算机应用,2021,(S1):1-6.
REN H,WANG X G.Review of attention mechanism[J].Journ-al of Computer Applications,2021,(S1):1-6.
[15] Jun Z,Nan C.A New Deep Transfer Learning Methodfor Bearing Fault Diagnosis Under Different Working Condition[J].IEEE Sensors Journal,2021,1(08):8394-8402.
[16] PAN S J, YANG Q. A survey on transfer learning [J].IEEE Transactions on Knowledge and Data Engineering,2010, 22(10): 1345-1359.
[17] 美国凯斯西储大学电机滚动轴承振动信号数据集[EB/OL].
[2021-07].https://csegroups.case.edu/bearingdatacenter.