One-dimensional Multi-scale Convolution Neural Network and its Application in Rolling Bearing Fault Diagnosis
-
摘要: 为了有效利用来自实际生产中监测系统的海量数据, 并结合一维卷积网络在处理一维数据的优势, 提出一种端到端的一维多尺度卷积神经网络滚动轴承故障诊断方法。首先使用两个一维卷积层和池化层将输入振动信号的长度缩减并增加通道数, 然后利用多尺度并行一维卷积核对上层输出特征进行不同尺度上的反复提取和重构, 最后将提取到的特征输入到一个全连接层进行故障分类。为验证算法的有效性, 通过对滚动轴承不同工况、不同训练样本以及与支持向量机、BP神经网络和循环神经网络等算法对比分析。结果表明提出的模型及方法具有较好的识别效果, 滚动轴承故障诊断正确率达到99.78%。Abstract: In order to use the massive data from the monitoring system in actual production effectively, and combine the advantages of one-dimensional convolution network in processing one-dimensional data, a new rolling bearing fault diagnosis method based on end-to-end one-dimension multi-scale convolution neural network is proposed. Firstly, two one-dimensional convolution layers and pooling layers are used to reduce the length of the input vibration signal and increase the number of channels. Then, multi-scale parallel one-dimensional convolution check is used to extract and reconstruct the output features on different scales repeatedly. Finally, the extracted features are input to a full connection layer for fault classification. In order to verify the effectiveness of the method, by comparing and analyzing the different working conditions, different training samples of rolling bearing and other algorithms such as support vector machine, BP neural network and cyclic neural network, the simulation results show that the proposed model and method has better recognition effect, and the accuracy of rolling bearing fault diagnosis reaches 99.78%.
-
Key words:
- rolling bearing /
- fault diagnosis /
- one-dimensional convolution network /
- deep learning
-
表 1 各分支中卷积核个数
分支 1×1卷积核个数 3×1卷积核个数 5×1卷积核个数 Branch A 64 - - Branch B 96 128 - Branch C 16 - 32 Branch D 32 - - 表 2 神经网络中数据尺寸变化
层名称 数据尺寸 Input 6 000×1 1_Conv1D 3 000×64 2_Conv1D 1 500×128 Max pooling 500×128 1d-MCL 1 500×256 1d-MCL 2 500×256 1d-MCL 3 500×256 Global Average Pooling 256×1 Dropout 256×1 FC/Output 10×1 表 3 试验数据集组成
故障类型 电机负载/hp 故障尺寸/mm 标签编号 正常 0/1/2/3 - 0 0.177 8 1 滚珠故障 0/1/2/3 0.355 6 2 0.533 4 3 0.177 8 4 内圈故障 0/1/2/3 0.355 6 5 0.533 4 6 0.177 8 7 外圈故障 0/1/2/3 0.355 6 8 0.533 4 9 表 4 不同训练集与测试集比例对诊断结果的影响
训练集与测试集比例 诊断准确率/% 9∶1 99.98 8∶2 99.94 7∶3 99.78 6∶4 99.65 5∶5 99.43 表 5 不同模型的滚动轴承故障诊断结果
数据集 1D-MCNN诊断准确率/% SVM诊断准确率/% BP网络诊断准确率/% RNN诊断准确率/% 0 hp 99.91 83.33 81.18 85.29 1 hp 99.82 95.61 82.94 98.53 2 hp 99.78 86.84 76.18 98.82 3 hp 99.91 93.86 87.06 99.41 -
[1] 袁文军, 刘飞, 王晓峰, 等. 基于深度自编码网络的轴承故障诊断[J]. 噪声与振动控制, 2018, 38(5): 208-214 doi: 10.3969/j.issn.1006-1355.2018.05.037YUAN W J, LIU F, WANG X F, et al. Bearing diagnosis based on deep neural network of auto-encoder[J]. Noise and Vibration Control, 2018, 38(5): 208-214 (in Chinese) doi: 10.3969/j.issn.1006-1355.2018.05.037 [2] HINTON G E, SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786): 504-507 doi: 10.1126/science.1127647 [3] UNAL M, ONAT M, DEMETGUL M, et al. Fault diagnosis of rolling bearings using a genetic algorithm optimized neural network[J]. Measurement, 2014, 58: 187-196 doi: 10.1016/j.measurement.2014.08.041 [4] 陈文戈, 赵学智. 轴承故障的自适应小波神经网络分类[J]. 轴承, 2009(3): 37-40 https://www.cnki.com.cn/Article/CJFDTOTAL-CUCW200903014.htmCHEN W G, ZHAO X Z. Fault classification of bearing using adaptive wavelet-based neural network[J]. Bearing, 2009(3): 37-40 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-CUCW200903014.htm [5] 黄竞楠, 王少红, 马超. 基于SVD-EEMD和BP神经网络的滚动轴承故障诊断[J]. 北京信息科技大学学报, 2019, 34(2): 69-74 https://www.cnki.com.cn/Article/CJFDTOTAL-BJGY201902014.htmHUANG J N, WANG S H, MA C. Fault diagnosis of rolling bearing based on SVD-EEMD and BP neural network[J]. Journal of Beijing Information Science & Technology University, 2019, 34(2): 69-74 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-BJGY201902014.htm [6] ZHANG J J, SONG Y X, QU Y. A Time series analysis and neural network based scheme for fault diagnosis of transformers[J]. Applied Mechanics and Materials, 2015, 742: 412-418 doi: 10.4028/www.scientific.net/AMM.742.412 [7] WANG X Q, LI Y F, RUI T, et al. Bearing fault diagnosis method based on Hilbert envelope spectrum and deep belief network[J]. Journal of Vibration Engineering, 2015, 17(3): 1295-1308 [8] KHAJAVI M N, KESHTAN M N. Intelligent fault classification of rolling bearings using neural network and discrete wavelet transform[J]. Journal of Vibration Engineering, 2014, 16(2): 761-769 http://www.jvejournals.com/article/14815/pdf [9] TAMILSELVAN P, WANG P F. Failure diagnosis using deep belief learning based health state classification[J]. Reliability Engineering & System Safety, 2013, 115: 124-135 [10] 庄雨璇, 李奇, 杨冰如, 等. 基于LSTM的轴承故障诊断端到端方法[J]. 噪声与振动控制, 2019, 39(6): 187-193 https://www.cnki.com.cn/Article/CJFDTOTAL-ZSZK201906034.htmZHUANG Y X, LI Q, YANG B R, et al. An end-to-end approach for bearing fault diagnosis based on LSTM[J]. Noise and Vibration Control, 2019, 39(6): 187-193 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZSZK201906034.htm [11] 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): 49-56 https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201521007.htmLEI Y G, JIA F, ZHOU X, et al. A deep learning-based method for machinery health monitoring with big data[J]. Journal of Mechanical Engineering, 2015, 51(21): 49-56 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXXB201521007.htm [12] 曲建岭, 余路, 袁涛, 等. 基于一维卷积神经网络的滚动轴承自适应故障诊断算法[J]. 仪器仪表学报, 2018, 39(7): 134-143 https://www.cnki.com.cn/Article/CJFDTOTAL-YQXB201807017.htmQU J L, YU L, YUAN T, et al. Adaptive fault diagnosis algorithm for rolling bearings based on one-dimensional convolutional neural network[J]. Chinese Journal of Scientific Instrument, 2018, 39(7): 134-143 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YQXB201807017.htm [13] ZHANG W, PENG G L, LI C H, et al. A new deep learning model for fault diagnosis with good anti-noise and domain adaptation ability on raw vibration signals[J]. Sensors, 2017, 17(2): 425 doi: 10.3390/s17020425 [14] WANG F A, JIANG H K, SHAO H D, et al. An adaptive deep convolutional neural network for rolling bearing fault diagnosis[J]. Measurement Science and Technology, 2017, 28(9): 095005 doi: 10.1088/1361-6501/aa6e22 [15] 吴春志, 江鹏程, 冯辅周, 等. 基于一维卷积神经网络的齿轮箱故障诊断[J]. 振动与冲击, 2018, 37(22): 51-56 https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201822008.htmWU C Z, JIANG P C, FENG F Z, et al. Faults diagnosis method for gearboxes based on a 1-D convolutional neural network[J]. Journal of Vibration and Shock, 2018, 37(22): 51-56 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201822008.htm [16] DING Y H, JIANG M P. A multi-scale convolutional auto-encoder and its application in fault diagnosis of rolling bearings[J]. Journal of Southeast University (English Edition), 2019, 35(4): 417-42 [17] 董靖川, 徐明达, 王太勇, 等. 分布式卷积神经网络在刀具磨损量预测中的应用[J]. 机械科学与技术, 2020, 39(3): 329-335 doi: 10.13433/j.cnki.1003-8728.20190131DONG J C, XU M D, WANG T Y, et al. Application of distributed convolutional neural network in wear prediction of Tool[J]. Mechanical Science and Technology for Aerospace Engineering, 2020, 39(3): 329-335 (in Chinese) doi: 10.13433/j.cnki.1003-8728.20190131 [18] 吴斌, 王敏杰, 康晶, 等. 滚动轴承故障振动信号特征与诊断方法[J]. 大连理工大学学报, 2013, 53(1): 76-81 https://www.cnki.com.cn/Article/CJFDTOTAL-DLLG201301015.htmWU B, WANG M J, KANG J, et al. Fault vibration signal feature of rolling bearing and its diagnosis method[J]. Journal of Dalian University of Technology, 2013, 53(1): 76-81 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-DLLG201301015.htm [19] SMITH W A, RANDALL R B. Rolling element bearing diagnostics using the case western Reserve University data: A benchmark study[J]. Mechanical Systems and Signal Processing, 2015, 64-65: 100-131 https://www.sciencedirect.com/science/article/pii/S0888327015002034 [20] VAN DER MAATEN L, HINTON G. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9(11): 2579-2605