针对小样本强噪声环境下，传统深度学习模型抗噪性差，模型训练不充分等问题，提出一种基于自适应最大二阶循环平稳盲解卷积(Adaptive Maximum Second -order Cyclostationarity Blind Deconvolution, ACYCBD)结合马尔可夫变迁场(Markov Transition Field, MTF)与MobileViT的滚动轴承故障诊断方法。首先，通过参数自适应的CYCBD算法增强强噪声背景下轴承故障的冲击信号，降低强背景噪声的影响，然后，采用MTF将预处理后的一维轴承振动信号转变为具有时间关联性的二维特征图像，最后将MTF图像输入MobileViT网络中进行训练，得到故障诊断结果，运用东南大学齿轮箱数据集和沈阳工业大学实验室滚动轴承数据集验证所提方法在小样本强噪声条件下的故障识别准确率，结果表明，在小样本强噪声条件下，ACYCBD处理后的数据，训练的模型具有更高的准确率，相较于其他数据预处理方法最大相关峭度解卷积、VMD、集合经验模态分解准确率分别提高了1.73、1.99、2.2个百分点，利用MTF进行模态转换后相较于格拉姆角场、连续小波变换、RP准确率分别高出了2.59、3.12、2.72个百分点；与其他深度学习模型进行对比，所提方法在上述条件下有着更高的抗干扰能力和泛化性能。
Abstract
Aiming at the problems of poor noise immunity and insufficient model training of traditional deep learning models in a small-sample and strong noise environment, a method based on Adaptive Maximum Second -order Cyclostationarity Blind Deconvolution (ACYCBD) combining Markov (ACYCBD) combined with Markov Transition Field (MTF) and MobileViT for rolling bearing fault diagnosis. Firstly, the impact signal of bearing faults under strong noise background is enhanced by parameter-adaptive CYCBD algorithm to reduce the influence of strong background noise, then, MTF is used to transform the preprocessed one-dimensional bearing vibration signal into a two-dimensional feature image with temporal correlation, and finally, the MTF image is inputted into the MobileViT network for training to get the fault diagnosis results, which is applied to the Southeast University Gearbox dataset and Shenyang University of Technology laboratory rolling bearing dataset to verify the fault identification accuracy of the proposed method in small sample strong noise conditions, the results show that, in the small sample strong noise conditions, ACYCBD processed data, the trained model has a higher accuracy, compared to maximum correlated kurtosis deconvolution, variational mode decomposition, ensemble empirical mode decompositionaccuracy increased by 1.73, 1.99, 2.2 After using MTF for modal transformation, the accuracy is 2.59, 3.12 and 2.72 percentage points higher than that of Gramian angular field, continuous wavelet transform and RP, respectively; comparing with other deep learning models, the method proposed in this paper has higher anti-interference ability and generalization performance under the above conditions.
关键词
滚动轴承 /
最大二阶循环平稳盲解卷积 /
马尔可夫变迁场 /
多头自注意力机制
{{custom_keyword}} /
Key words
Rolling bearings /
Maximum second-order blind inverse convolution /
Markov transformation field /
Multi-head Self-Attention Machanism
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 陈子旭，朱振杰，卢国梁．一种新的图谱域滚动轴承早期故障检测与识别方法[J]．振动与冲击，2022，41(6)：51-59.
CHEN Zixu, ZHU Zhenjie, LU Guoliang. Novel early fault detection and diagnosis for rolling element bearings in graph spectrum domain[J]. Journal of Vibration and Shock, 2022, 41(6): 51-59.
[2] 殿勇，薛红涛，刘炳晨．基于OSSD-EMOMEDA的轮毂电机轴承故障特征提取方法[J]．中国电机工程学报：2023，43(20)：7759-7743．
DIAN Yong XUE Hongtao, LIU Bingchen. Feature Extraction Method based on Optimized SSD and Enhance MOMEDA for Bearing Faults of In-wheel Motor[J]. Proceedings of the CSEE, 2023, 43(20): 7759-7743.
[3] Huang N E, Shen Z, Long S R, et al. The empirical mode decomposition and the hilbert Spectrum for nonlinear and non-stationary time series analysis[J]. Proceedings of Royal Society of London, 1998, 454(1971): 903-995.
[4] Chen H G, Chen P, Chen W H, et al. Wind turbine gearbox fault diagnosis based on improved EEMD and Hilbert square demodulation[J]. Applied Sciences-Basel, 2017, 7(2):1-15.
[5] Ma J, Wu J D, Wang Xiaodong. A hybrid fault diagnosis method based on singular value difference spectrum denoising and local mean decomposition for rolling bearing[J]. Journal of Low Frequency Noise Vibration and Active Control, 2018, 37(4):928-95.
[6] 陈是扦，彭志科，周鹏．信号分解及其在机械故障诊断中的应用研究综述[J]．机械工程学报，2020，56(17)：91-107．
CHEN Shiqian, PENG Zhike, ZHOU Peng. Review of Signal Decomposition Theory and Its Applications in Machine Fault Diagnosis [J]. Journal of Mechanical Engineering, 2020, 56(17): 91-107.
[7] Cheng Y, Chen B Y, Zhang W H. Adaptive multipoint optimal minimum entropy deconvolution adjusted and application to fault diagnosis of rolling element bearings[J]. IEEE Sensors Journal, 2019, 19(24): 12153-12164.
[8] 周涛，赵明，郭栋，等．基于信号子空间的新型盲解卷积方法[J]．振动与冲击，2022，41(3)：139-147．
ZHOU Tao, ZHAO Ming, GUO Dong, et al. A new blind deconvolution method based on signal subspace[J]. Journal of Vibration and Shock, 2023, 2562(1).
[9] Tang G, Wang X, He Y. Diagnosis of compound faults ofrolling bearings through adaptive maximum correlated kurtosis deconvolution[J]. Journal of Mechanical Science and Technology, 2016, 30(1):43-54.
[10] Hong L, Liu X Z, UO H. Compound fault diagnosis of rotating machinery based on adaptive maximum correlated kurtosis deconvolution and customized multiwavelet transform[J]. Measurement Science and Technology, 2018, 29 (11):115007.
[11] Miao Y, Zhao M, Lin J, et al. Application of an improved maximum correlated kurtosis deconvolution method for fault diagnosis of rolling element bearings[J]. Mechanical Systems and Signal Processing, 2017, 92:173-195.
[12] Buzzoni M, Antoni J, D’ELIA G. Blind deconvolution based on cyclostationarity maximization and its application to fault identification[J]. Journal of Sound and Vibration, 2018, 432: 569-601．
[13] 古莹奎，吴宽，李成．基于格拉姆角场和迁移深度残差神经网络的滚动轴承故障诊断[J]．振动与冲击，2022，41(21)：228-237．
GU Yingkui, WU Kuan, LI Cheng. Rolling bearing fault diagnosis based on Gram angle field and transfer deep residual neural network[J]. Journal of Vibration and Shock, 2022, 41(21): 228-237.
[14] 张龙，胡燕青，赵丽娟，等．采用递归图编码技术与残差网络的滚动轴承故障诊断[J]．西安交通大学学报，2023，57(02)：110-120．
ZHANG Long, HU Yanqing, ZHAO Lijuan, et al. Fault Diagnosis of Rolling Bearings Using Recurrence Plot Coding Technique and Residual Network[J]. Joernal of Xi’an Jiaotong University, 2023, 57(02): 110-120.
[15] 雷春丽，薛林林，夏奔锋，等．采用马尔科夫转移场和图注意力网络的滚动轴承故障诊断方法[J]．振动工程学报，2023，36(05)：1179-1187．
LEI Chunli, XUE Linlin, XIA Mangfeng, et al. Rolling Bearing Fault Diagnosis Method Based on Markov Transition Field and Graph Attention Network[J]. Journal of Vibration Engineering, 2023, 36(05): 1179-1187.
[16] 赵志宏，吴冬冬，窦广鉴，等. 一种基于CNN-BiGRU孪生网络的轴承故障诊断方法[J]. 振动与冲击，2023，42(06):166-171+211.
ZHAO Zhihong,WU Dongdong,DOU Guangjian,et al.Bearing fault diagnosis method based on a CNN-BiGRU siamese network[J]. Journal of Vibration and Shock, 2023，42(6):166-171.
[17] 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.
[18] 雷春丽，夏奔锋，薛林林，等. 优化深度残差网络及其在强噪声环境下滚动轴承故障诊断中的应用[J]. 振动工程学报，2023，36(06):1754-1763.
LEI Chunli,XIA Benfeng,XUE Linlin,et al.Optimized deep residual network and its application in fault diagnosis of rolling bearing under the strong noise condition[J]. Journal of Vibration Engineering,2023，36(6):1754-1763.
[19] Zhang W, Li C, Peng G A deep convolutional neural network with new training methods for bearing fault diagnosis under noisy environment and different working load[J]. Mechanical Systems and Signal Processing,2018, 100: 439–453.
[20] Liang M X, Cao P, Tang J. Rolling bearing fault diagnosis based on feature fusion with parallel convolutional neural network[J]. The International Journal of Advanced Manufacturing Technology, 2021, 112: 819-831.
[21] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[J]. Advances in neural information processing systems, 2017, 30(02): 18-26.
[22] 项伟，刘淑杰，李宏坤，等．参数自适应CYCBD的滚动轴承复合故障特征提取[J]．航空动力学报，2023，37(12)：1-9.
XIANG Wei, LIU Shujie, LI Hongkun, et al. Compound fault feature extraction of rolling bearing based on parameters adaptive CYCBD [J]. Journal of Aerospace Power, 2023, 37(12): 1-9.
[23] 黄包裕，张永祥，赵磊．基于布谷鸟搜索算法和最大二阶循环平稳盲解卷积的滚动轴承故障诊断方法[J]．机械工程学报，2021，57(09)：99-107．
HUANG Baoyu, ZHANG Yongxiang, ZHAO Lei. Research on Fault Diagnosis Method of Rolling Bearings Based on Cuckoo Search Algorithm and Maximum Second Order Cyclostationary Blind Deconvolution[J]. Journal of Sound and Vibration, 2018, 432: 569-601．
[24] Wang Z G, Oates T. Encoding time series as imag‑es for visual inspection and classification using tiled con‑volutional neural networks[C]. Workshops at theTwenty-ninth AAAI Conference on Artificial Intelli‑gence, Austin, 2015: 1-7.
[25] 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(9)：151-158．
LEI Chunli, XIA Mangfeng, XUE Linlin, et al. Rolling bearing fault diagnosis method based on MTF-CNN[J]. Journal of Vibration and Shock, 2022, 41(9): 151-158.
[26] Jing L, Huiyang C, Weimin Z. Improved MobileViT: A More Efficient Light-weight Convolution and Vision Transformer Hybrid Model[J]. Journal of Physics: Conference Series, 2023, 2562(1): 48-57.
[27] 刘俊锋，俞翔，万海波，等．基于MFMD和Transformer-CNN的滚动轴承故障诊断方法[J]．航空动力学报，2023，38(06)：1446-1456．
LIU Junfeng, YU Xiang, WAN Haibo, et al. Fault diagnosis method of rolling bearing using MFMD and Transformer-CNN[J]. Journal of Aerospace Power, 2023, 38(06): 1446-1456.
[28] 刘茜，易诗，李立，等．基于轻量级CNN-Transformer混合网络的梯田图像语义分割[J]．农业工程学报，2023，39(13)：171-181．
LIU Qian, YI Shi, LI Li, et al. ISemantic segmentation of terrace image regions based on lightweight CNN-Transformer hybrid networks. [J]. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39(13): 171-181.
[29] Shao S, McAleer S, Yan R, et al. Highly accurate machine fault diagnosis using deep transfer learning[J]. IEEE Transactions on Industrial Informatics, 2018, 15(4): 2446-2455.
[30] Dong Y, Li Y, Zheng H, et al. A new dynamic model and transfer learning based intelligent fault diagnosis framework for rolling element bearings race faults: Solving the small sample problem[J]. ISA transactions, 2022, 121:327-348.
[31] Zhang X, He C, Lu Y, et al. Fault diagnosis for small samples based on attention mechanism[J]. Measurement, 2022, 187: 110242.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(3705 KB)
