Rolling bearing fault diagnosis algorithm based on improved alternating transfer learning

PDF(4318 KB)

Journal of Vibration and Shock ›› 2024, Vol. 43 ›› Issue (5) : 239-249.

WANG Peng,LI Danqing,WANG Heng

Author information +

History +

Abstract

For the problems of large differences in feature distribution between monitoring signals of different models of rolling bearings and small samples of fault data, which lead to low accuracy of bearing faults, a rolling bearing fault diagnosis algorithm based on improved alternating migration learning is proposed in this paper. In order to give full play to the excellent feature extraction ability of convolutional neural networks (CNN) for two-dimensional data, firstly, the one-dimensional vibration signal is transformed into a two-dimensional image and input to a deep convolutional neural network for learning; secondly, in order to reduce the difference of feature distribution between the source and target domain data, an improved alternately transfer learning (IATL) is proposed to calculate the CORAL loss function and maximum mean discrepancy (MMD) loss function between domains alternately, and back propagate to update the weights and bias parameters of each layer of the network, in order to achieve migration adaptation of bearing characteristics under variable operating conditions, across bearing types and small failure sample conditions; Finally, the Softmax function is used in the fully connected layer for fault diagnosis of the target domain data. In order to verify the effectiveness of the proposed algorithm, a migration experiment is conducted using rolling bearing dataset from Case western reserve university (CWRU). The results show that the algorithm effectively reduces the difference of feature distribution between domain data and has higher fault classification accuracy when compared with algorithms such as calculating only CORAL loss function and MMD loss function.

Key words

rolling bearing / fault diagnosis / transfer learning / loss function / deep convolutional neural networks

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

WANG Peng,LI Danqing,WANG Heng. Rolling bearing fault diagnosis algorithm based on improved alternating transfer learning[J]. Journal of Vibration and Shock, 2024, 43(5): 239-249

References

[1] Zhu W, Ni G, Cao Y, et al. Research on a rolling bearing health monitoring algorithm oriented to industrial big data[J]. Measurement, 2021, 185: 110044. [2] 李军宁，罗文广，陈武阁．面向振动信号的滚动轴承故障诊断算法综述[J].西安工业大学学报2022, 42(02): 105-122. Li J N, Luo W G, Chen W G. Overview of Algorithms for Rolling Bearing Fault Diagnosis Based on Vibration Signal. [J]. Journal of Xi’an Technological University, 2022, 42(02):105-122. [3] Li J, Liu Y, Li Q. Intelligent fault diagnosis of rolling bearings under imbalanced data conditions using attention-based deep learning method[J]. Measurement, 2022, 189: 110500. [4] Chen X, Shu G, Zhang K, et al. A fault characteristics extraction method for rolling bearing with variable rotational speed using adaptive time-varying comb filtering and order tracking[J]. Journal of Mechanical Science and Technology, 2022, 36(3): 1171-1182. [5] Wang C, Jiang W, Yang X, et al. RUL Prediction of Rolling Bearings Based on a DCAE and CNN[J]. Applied Sciences, 2021, 11(23): 11516. [6] 雷春丽，夏奔锋，薛林林，等. 基于MTF-CNN的滚动轴承故障诊断方法[J]. 振动与冲击, 2022,41(09):151-158. LEI C L, JIAO M X, MA S Z. Fault Diagnosis Method of Small Sample Rolling Bearing under Variable Working Conditions Based on MTF-SPCNN[J]. Journal of Vibration and Shock, 2022,41(09):151-158. [7] Xu Q, Zhu B, Huo H, et al. Fault diagnosis of rolling bearing based on online transfer convolutional neural network[J]. Applied Acoustics, 2022, 192: 108703. [8] Zheng Z, Fu J, Lu C, et al. Research on rolling bearing fault diagnosis of small dataset based on a new optimal transfer learning network[J]. Measurement, 2021, 177: 109285. [9] 黄迅迪，庞雄文.基于深度学习的智能设备故障诊断研究综述[J].计算机科学，2023，1-12. HUANG X D, PANG X W. Review of Intelligent Device Fault Diagnosis Based on Deep Learning[J]. Computer Science，2023，1-12. [10] Zhou J, Yang X, Li J. Deep Residual Network Combined with Transfer Learning Based Fault Diagnosis for Rolling Bearing[J]. Applied Sciences, 2022, 12(15): 7810. [11] Wang L, Liu H, Pan Z, et al. Long Short-Term Memory Neural Network with Transfer Learning and Ensemble Learning for Remaining Useful Life Prediction[J]. Sensors, 2022, 22(15): 5744. [12] 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]. 机械工程学报, 2023, 59 (3): 76-85. SHAO H D, XIAO Y M, YAN S. Simulation Data-driven Enhanced Unsupervised Domain Adaptation for Bearing Fault Diagnosis. [J]. Journal Of Mechanical Engineering, 2023, 59 (3): 76-85. [13] 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(03): 159-164+173. SHI J W, HOU L Q. Bearing fault diagnosis based on 1D CNN attention gated recurrent network and transfer learning[J]. Journal of Vibration and Shock, 2023, 42(03): 159-164+173. [14] LONG M，CAO Y，WANG J，et al． Learning transferable features with deep adaptation networks[J]. Proceedings of the 32nd International Conference on International Conference on Machine Learning，2015, 37( 7) : 97-105. [15] Yang B, Lei Y, Jia F, 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. [16] Huo C, Jiang Q, Shen Y, et al. New transfer learning fault diagnosis method of rolling bearing based on ADC-CNN and LATL under variable conditions[J]. Measurement, 2022, 188: 110587. [17] Deng M, Deng A, Shi Y, et al. Correlation regularized conditional adversarial adaptation for multi-target-domain fault diagnosis[J]. IEEE Transactions on Industrial Informatics, 2022, 18(12): 8692-8702.