基于多维深度特征融合与改进麻雀搜索卷积神经网络的滚动轴承故障声发射诊断

PDF(3153 KB)

振动与冲击 ›› 2023, Vol. 42 ›› Issue (7) : 65-76.

论文

魏巍1，2，王之海1，2，柳小勤1，2，冯正江1，2，李佳慧1，2

作者信息 +

AE Diagnosis of rolling bearing faults based on MDFF and ISSA

WEI Wei1,2, WANG Zhihai1,2, LIU Xiaoqin1,2, FENG Zhengjiang1,2, LI Jiahui1,2

Author information +

文章历史 +

摘要

针对滚动轴承早期、复合故障难以准确诊断与智能诊断模型超参数确定严重依赖专家先验知识问题，提出一种基于多维深度特征融合(MDFF)与改进麻雀搜索算法(ISSA)的滚动轴承故障声发射诊断方法。首先，用一维卷积与线性瓶颈反向残差二维卷积神经网络构建多输入卷积神经网络（CNN）结构的诊断模型，模型输入为滚动轴承声发射信号及其小波时频图，提出基于布伦纳梯度和信噪比的质量指标，在108种小波基中筛选出最佳时频图以提升输入数据质量。接着，采用特征金字塔网络将模型的一、二维低层与高层特征融合，建立深度融合的诊断模型。然后，将交叉混沌映射、自适应权重及融合的随机游走策略引入麻雀搜索算法中，以自适应获取MDFFCNN最优超参数。试验表明，对比近期多个主流智能诊断算法，所提方法可避免人工选择诊断模型超参数，对滚动轴承早期尤其复合故障具有更高的诊断精度和稳定性，模型诊断过程的智能化水平得到了进一步提高。

Abstract

Aiming at the problem that the early and complex faults of rolling bearings are difficult to accurately diagnose and the determination of hyperparameters of the intelligent diagnosis model depends heavily on the prior knowledge of experts, an acoustic emission diagnosis method for rolling bearing faults based on Multidimensional Deep Feature Fusion (MDFF) and Improved Sparrow Search Algorithm (ISSA) is proposed. First, a multi-input convolutional neural network (CNN) diagnosis model is constructed using one-dimensional convolution and linear bottleneck inverse residual two-dimensional convolutional neural network. The input of the model is the rolling bearing acoustic emission signal and its wavelet time-frequency diagram. A quality index based on Brenner gradient and signal-to-noise ratio is proposed, and the best time-frequency image is selected from 108 wavelet bases to improve the quality of input data. Then, the feature pyramid network is used to fuse the one- and two-dimensional low-level and high-level features of the model to establish a deep fusion diagnostic model. Then, the random walk strategy of cross chaos mapping and adaptive weighting and fusion is introduced into the sparrow search algorithm to adaptively obtain the optimal hyperparameters of MDFFCNN. Experiments have shown that, compared with multiple mainstream intelligent diagnosis algorithms in the recent past, the proposed method can avoid manual selection of hyperparameters of the diagnosis model, and has higher diagnosis accuracy and stability for early-stage especially compound faults of rolling bearings, and the intelligent level of the model diagnosis process has been further improved.

导出引用

魏巍1，2，王之海1，2，柳小勤1，2，冯正江1，2，李佳慧1，2. 基于多维深度特征融合与改进麻雀搜索卷积神经网络的滚动轴承故障声发射诊断[J]. 振动与冲击, 2023, 42(7): 65-76

WEI Wei1,2, WANG Zhihai1,2, LIU Xiaoqin1,2, FENG Zhengjiang1,2, LI Jiahui1,2. AE Diagnosis of rolling bearing faults based on MDFF and ISSA[J]. Journal of Vibration and Shock, 2023, 42(7): 65-76

参考文献

[1] 雷亚国,贾峰,孔德同,等.大数据下机械智能故障诊断的机遇与挑战[J].机械工程学报,2018,54(05):94-104.
LEI Yaguo, JIA Feng, KONG Detong, et al. Opportunities and Challenges of Machinery Intelligent Fault Diagnosis in Big Data Era[J]. Journal of Mechanical Engineering, 2018, 54(05): 94-104.
[2] VAN H B, He D, Qu Y. On the use of spectral averaging of acoustic emission signals for bearing fault diagnostics[J]. Journal of Vibration and Acoustics, 2014, 136(6).
[3] 常丽,赵若君.基于声发射的风机轴承故障诊断研究[J].仪表技术与传感器,2017(09):109-113.
CHANG Li, ZHAO Ruojun. Research of Wind Turbines Bearing Fault Diagnosis [J]. Instrument Technique and Sensor, 2017 (09):109-113.
Based on Acoustic Emission
[4] XU T, FENG Z G. Tool wear identifying based on EMD and SVM with AE sensor[C]//2009 9th International Conference on Electronic Measurement & Instruments. IEEE, 2009: 2-948-2-952.
[5] ZHANG Z, YANG G, HU K. Prediction of Fatigue Crack Growth in Gas Turbine Engine Blades Using Acoustic Emission[J]. Sensors, 2018, 18(5):1321.
[6] ACHMAD W, ERIC Y K, JONG D S, et al. Fault diagnosis of low speed bearing based on relevance vector machine and support vector machine. Expert Systems with Applications: An International Journal archive, 2009,3(36): 7252-7261.
[7] SUN S, GUO Y, GAO Y. Experimental investigation on double impulses phenomenon of outer race spalled rolling element bearings based on acoustic emission signals[C]//2015 IEEE International Conference on Information and Automation. IEEE, 2015: 1931-1935.
[8] 郑胜,刘韬,刘畅,等. 基于级联过完备字典稀疏表征的滚动轴承复合故障诊断方法[J]. 振动与冲击,2021, 40(10):7.
ZHENG Sheng, LIU Tao, LIU Chang, et al. Compound faults diagnosis method of rolling bearing based on sparse representation of cascaded over complete dictionary[J]. Journal of Vibration and Shock,2021, 40(10):7.
[9] LIU F, ZHOU S, XIA C, et al. Optimization of fatigue life distribution model and establishment of probabilistic S–N curves for a 165 ksi grade super high strength drill pipe steel[J]. Journal of Petroleum Science & Engineering, 2016:527-532.
[10] 张晓涛,唐力伟,王平,等. 最小周期相关熵解卷积结合窄带解调的轴承复合故障诊断研究[J]. 振动工程学报, 2015, 000(004):666-672.
ZHANG Xiaotao, TANG Liwei, WANG Ping, et al. Bearing with minimum period correlation entropy deconvolution combined with narrow band demodulation research on compound fault diagnosis[J]. Journal of Vibration Engineering, 2015, 000 (004): 666-672.
[11] APPANA D K , ALECANDER P , JONG M K. Reliable fault diagnosis of bearings with varying rotational speeds using envelope spectrum and convolution neural networks[J]. Soft Computing, 2018:1-11.
[12] ISLAM M, KIM J M. Automated bearing fault diagnosis scheme using 2D representation of wavelet packet transform and deep convolutional neural network[J]. Computers in Industry, 2019, 106:142-153.
[13] 施杰,伍星,刘韬. 采用HHT算法与卷积神经网络的轴承复合故障诊断[J]. 农业工程学报, 2020, 36(4):34-43.
SHI Jie, WU Xing, LIU Tao. Compound fault diagnosis of bearing based on HHT algorithm and convolutional neural network[J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(4):34-43.
[14] SAUFI S R, AHMAD Z A B, LEONG M S, et al. Low- Speed Bearing Fault Diagnosis Based on ArSSAE Model Using Acoustic Emission and Vibration Signals[J]. IEEE Access, 2019, 7:46885-46897.
[15] 李涛,段礼祥,张东宁,等. 自适应卷积神经网络在旋转机械故障诊断中的应用[J]. 振动与冲击, 2020, 39(16):9.
LI Tao, DUAN Lixiang, ZHANG Dongnin, et al. Application of adaptive convolutional neural network in rotating machinery fault diagnosis[J]. Journal of Vibration and Shock, 2020, 39(16):9.
[16] WANG F, JIANG H, Shao H, et al. An adaptive deep convolutional neural network for rolling bearing fault diagnosis[J]. Measurement Science and Technology, 2017, 28(9):095005.
[17] XUE J, SHEN B. A novel swarm intelligence optimization approach: sparrow search algorithm[J]. Systems Science & Control Engineering, 2020, 8(1) : 22-34.
[18] SUN W, DENG A, DENG M, et al. Multi-view feature fusion for rolling bearing fault diagnosis using random forest and autoencoder[J]. Journal of Southeast University (English Edition),2019,35(03):302-309.
[19] HOWARD A, SANDLER M, CHU G, et al. Searching for mobilenetv3[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2019: 1314-1324.
[20] Krizhevsky A, Hinton G. Convolutional deep belief networks on cifar-10[J]. Unpublished manuscript, 2010, 40(7): 1-9.
[21] LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2117-2125.
[22] SHEN Q, JIANG Y, SHEN Z, et al. Test case minimizing based on combination chaos genetic algorithm[J]. Telecommunications Science, 32(6): 93.
[23] 朱和贵,蒲宝明,朱志良,赵怡然,宋禹佳.二维Sine-Tent超混沌映射及其在图像加密中的应用[J].小型微型计算机系统,2019,40(07):1510-1518.
ZHU Hegui, PU Baoming, ZHU Zhiliang, et al. Two-dimensional Sine-Tent-based Hyper Chaotic Map and its Application in Image Encryption[J]. Journal of Chinese Computer Systems, 2019,40(07):1510-1518.
[24] 吴小龙,胡松,成卫. 基于改进鲸鱼优化算法的多目标信号配时优化[J]. 昆明理工大学学报：自然科学版, 2021, 46(1):8.
WU Xiaolong, HU Song, CHENG Wei. Multi － Objective Signal Timing Optimization Based on Improved Whale Optimization Algorithm[J]. Journal of Kunming University of Science and Technology( Natural Sciences) , 2021, 46(1):8.
[25] KINGMA D, BA J. Adam: A method for stochastic optimi- zation [J]. arXiv preprint arXiv:1412.6980, 2014: 273-297.
[26] 任学平,黄慧杰,王朝阁,等.改进的TQWT在滚动轴承早期故障诊断的应用[J].振动.测试与诊断,2020,40(02): 317-325+420.
REN Xueping, HUANG Huijie, WANG Chaoge, et al. Application of improved TQWT in early fault diagnosis of rolling bearings[J]. Journal of Vibration, Measurement& Diagnosis,2020,40(02): 317-325+420.
[27] XU K, LI S, WANG J, et al. A novel adaptive and fast deep convolutional neural network for bearing fault diagnosis under different working conditions[J]. Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering, 2019, 234(4):095440701986102.
[28] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770-778.
[29] 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): 147-154+168.
ZAN Tao, WANG Hui, LIU Zhihao, et al. A fault diagnosis model for rolling bearing based on a multi-input layer convolutional neural network[J]. Journal of Vibration and Shock, 2020,39 (12) :147-154+168.
[30] LAURENS V D M, HINTON G. Visualizing Data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9(2605):2579-2