基于多尺度特征融合残差神经网络的旋转机械故障诊断

PDF(2215 KB)

振动与冲击 ›› 2021, Vol. 40 ›› Issue (24) : 22-28.

论文

邓飞跃1,2,3，丁浩3，郝如江1,2

作者信息 +

Fault diagnosis of rotating machinery based on residual neural network with multi-scale feature fusion

DENG Feiyue1,2,3，DING Hao3，HAO Rujiang1,2

Author information +

文章历史 +

摘要

轴承、齿轮等旋转部件常在复杂工况下运行，环境噪声干扰大，导致故障特征微弱而难以准确诊断。基于此，本文提出一种新的多尺度特征融合残差块（Multi-scale feature fusion residual block, MSFFRB）设计方法，基于此构建了一维残差神经网络用于旋转机械故障诊断。该模型能够将不同尺度的网络卷积层级联在一起提取多尺度特征信息，在残差块内部实现了多尺度特征信息的有效融合，兼顾了残差网络跨层恒等映射与多尺度特征提取的优势，克服了传统卷积操作只能提取单一尺度特征信息的缺点。所构建的残差神经网络可以直接输入样本数据，不需要进行任何数据预处理，而且模型结构具有较高的灵活性，易于扩展。实验分析表明，所提网络可有效用于旋转机械的故障诊断，相比传统CNNs、ResNets、1D-LeNets、1D-AlexNets、MC-CNNs等5种当前常用网络，具有更好的抗噪性能，故障分类准确率更高，这为旋转机械故障诊断提供了一种新的途径。

Abstract

Aiming at the problem that weak fault feature of rotating machinery, such as rolling bearing and gear, is difficult to detect under strong background noise, this paper proposes a novel design method of multi-scale feature fusion residual block (MSFFRB), and a one-dimensional (1D) residual neural network is developed to diagnose fault of rotating machinery. In the proposed MSFFRB, some convolutional layers with different scales are cascaded together to extract multi-scale feature information, and the multi-scale feature information is effectively fused. It simultaneously takes into account the advantages of crossing layer identity mapping and the multi-scale feature extraction, and overcomes the disadvantage that traditional convolution layer with fixed scale only extracts single scale feature information. The proposed network can input data directly without any data preprocessing. Moreover, the architecture of network has high flexibility and is easy to further expand. The experimental results show that this method can be effectively used for fault diagnosis of rotating machinery. Compared to the traditional CNNs, ResNets, 1D-LeNets, 1D-AlexNets and MC-CNNs, the proposed method has better anti-noise performance and higher classification accuracy, which provides a new solution for rotating machinery fault diagnosis.

导出引用

邓飞跃1,2,3，丁浩3，郝如江1,2. 基于多尺度特征融合残差神经网络的旋转机械故障诊断[J]. 振动与冲击, 2021, 40(24): 22-28

DENG Feiyue1,2,3，DING Hao3，HAO Rujiang1,2. Fault diagnosis of rotating machinery based on residual neural network with multi-scale feature fusion[J]. Journal of Vibration and Shock, 2021, 40(24): 22-28

参考文献

[1] R.B. Sun, Z.B. Yang, Z. Zhai, et al. Sparse representation based on parametric impulsive dictionary design for bearing fault diagnosis[J]. Mechanical systems and signal processing, 2019, 122:737–753.
[2] P.W. Tse, Y.H. Peng, R. Yam, Wavelet Analysis and Envelope Detection For Rolling Element Bearing Fault Diagnosis-Their Effectiveness and Flexibilities[J]. Journal of Vibration and Acoustics, 2001, 123(3):303-310.
[3] 蔡剑华, 王先春, 胡惟文. 基于经验模态分解与小波阈值的MT信号去噪方法[J]. 石油地球物理勘探, 2013, 48(2):303-307
Cai Jianhua, Wang Xianchun, Hu Weiwen. A method for MT data denoising based on empirical mode decomposition and wavelet threshold[J]. Oil Geophysical Prospecting, 2013, 48(2):303-307
[4] C.F. Liu, L.D. Zhu, C.B. Ni. Chatter detection in milling process based on VMD and energy entropy[J]. Mechanical systems and signal processing, 2018, 105:169-182
[5] M. Kedadouche, M. Thomas, A Tahan. A comparative study between Empirical Wavelet Transforms and Empirical Mode Decomposition Methods: Application to bearing defect diagnosis[J]. Mechanical systems and signal processing, 2016, 81:88-107.
[6] M. Cerrada, G. Zurita, D. Cabrera, et al. Fault diagnosis in spur gears based on genetic algorithm and random forest[J]. Mechanical systems and signal processing, 2015, 71-72:87-103
[7] S. Dong, T. Luo, L. Zhong, et al. Fault diagnosis of bearing based on the kernel principal component analysis and optimized k-nearest neighbour model[J]. Journal of Low Frequency Noise Vibration & Active Control, 2017, 36(4):354-365
[8] 雷亚国, 贾峰, 周昕等. 基于深度学习理论的机械装备大数据健康监测方法[J]. 机械工程学报, 2015, 51(21)：50-56.
Lei Yaguo, Jia Feng, Zhou Xin et al. A Deep Learning-based Method for Machinery Health Monitoring with Big Data[J]. Journal of Mechanical Engineering, 2015, 51(21):50-56.
[9] H.D. Shao, H.K. Jiang, X.Q. Li, et al. Rolling bearing fault detection using continuous deep belief network with locally linear embedding[J]. Computers in Industry, 2018, 96:27-39
[10] Z. Xiang, X.N. Zhang, W.W. Zhang, et al. Fault diagnosis of rolling bearing under fluctuating speed and variable load based on TCO Spectrum and Stacking Auto-encoder[J]. Measurement, 2019, 138:162-174
[11] 昝涛, 王辉, 刘智豪等. 基于多输入层卷积神经网络的滚动轴承故障诊断模型[J]. 振动与冲击，2020，39(12):142-149.
Jiu Tao, Wang Hui, Liu Zhihao et al. A fault diagnosis model for rolling bearings based on a multi-input layer convolutional neural network[J]. Journal of vibration and shock,2020, 39（12）：142-149.
[12] D. T. Hoang, H. J. Kang. Rolling Element Bearing Fault Diagnosis using Convolutional Neural Network and Vibration Image.Cognitive Systems Research, 2019,53:42-50
[13] 鄢仁武, 林穿, 高硕勋等. 基于小波时频图和卷积神经网络的断路器故障诊断分析[J]. 振动与冲击，2020，39(10):199-205.
Yan Renwu, Lin Chuan, Gao Shuoxun et al. Fault diagnosis and analysis of circuit breaker based on wavelet time-frequency representations and convolution neural network[J]. Journal of Vibration and Shock, 2020, 39(10):199-205.
[14] M.H. Zhao,M. Kang. B.P. Tang, et al. Deep Residual Networks With DynamicallyWeighted Wavelet Coefficients for FaultDiagnosis of Planetary Gearboxes[J]. IEEE Transactions on Industrial Electronics, 2018, 65(5):4290-4300
[15] W. Zhang, C.H. Li, G.L. Peng, et al. A deep convolutional neural network with new trainingmethods for bearing fault diagnosis under noisy environmentand different working load[J]. Mechanical systems and signal processing, 2018, 100 439–453
[16] Zhao M , Kang M , Tang B , et al. Deep Residual Networks With Dynamically Weighted Wavelet Coefficients for Fault Diagnosis of Planetary Gearboxes[J]. IEEE Transactions on Industrial Electronics, 2018, 65(5):4290-4300.
[17] K. He, X. Zhang, S. Ren et al. Identity mappings in deep residual networks[J]. European Conference on Computer Vision(ECCV) 2016, Cham, Switzerland: Springer, 2016, 630–645.
[18] W.Y. Huang, J.S. Chen, Y. Yang, et al. An improved deep convolutional neural network with multi-scale information for bearing fault diagnosis[J]. Neurocomputing, 2019,359: 77–92.
[19] Z. Jinde, J. Zhanwei, P. Haiyang, Sigmoid-based refined composite multiscale fuzzy entropy and t-SNE based fault diagnosis approach for rolling bearing[J]. Measurement, 2018, 129 332–342.
[20] Feiyue Deng, Shaopu Yang, Guiji Tang, et al. Self adaptive multi-scale morphology AVG-Hat filter and its application to fault feature extraction for wheel bearing[J]. Measurement Science and Technology, 2017, 28(4):045011
[21] P. Li, F. Kong, Q. He, et al., Multiscale slope feature extraction for rotating machinery fault diagnosis using wavelet analysis[J], Measurement, 2013, 46 (1):497–505.
[22] 朱会杰, 王新晴, 芮挺等. 基于平移不变CNN的机械故障诊断研究[J]. 振动与冲击，2019，38(5):45-52.
Zhu Huijie, Wang Xinqing, Rui Ting, et al. Machinery fault diagnosis based on shift invariant CNN[J]. Journal of vibration and shock, 2019，38(5):45-52.
[23] F.J. Jia, Y.G. Lei, N. Lu, et al. Deep normalized convolutional neural network for imbalanced fault classification of machinery and its understanding via visualization[J]. Mechanical systems and signal processing, 2018, 110:349-367