|本期目录/Table of Contents|

[1]张 伟,张广帅,王连彪.基于CNN-GRU网络的轴承故障检测算法[J].工业仪表与自动化装置,2021,(06):88-91.[doi:10.19950/j.cnki.cn61-1121/th.2021.06.017]
 ZHANG Wei,ZHANG Guangshuai,WANG Lianbiao.Bearing fault detection algorithm based on CNN-GRU network[J].Industrial Instrumentation & Automation,2021,(06):88-91.[doi:10.19950/j.cnki.cn61-1121/th.2021.06.017]
点击复制

基于CNN-GRU网络的轴承故障检测算法

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2021年06期
页码:
88-91
栏目:
出版日期:
2021-12-15

文章信息/Info

Title:
Bearing fault detection algorithm based on CNN-GRU network
作者:
张 伟张广帅王连彪
(青岛科技大学 自动化与电子工程学院,山东 青岛 266100)
Author(s):
ZHANG Wei ZHANG Guangshuai WANG Lianbiao
(College of Automation and Electronic Engineering, Qingdao University of Science and Technology,Shandong Qingdao 266100,China)
关键词:
轴承故障诊断CNNGRUeLU
Keywords:
bearing fault diagnosis CNN GRU eLU
分类号:
TH117
DOI:
10.19950/j.cnki.cn61-1121/th.2021.06.017
文献标志码:
A
摘要:
传统的卷积神经网络(Convolutional Neural Networks , CNN)采用的激活函数sigmoid,tanh计算量大,容易出现梯度消失和过拟合现象,ReLU函数容易导致均值偏移,且在轴承故障诊断中,故障数据一般为时序信号。为了解决CNN网络中激活函数存在的问题,同时提高对故障数据的时序特征提取能力,将CNN与门控循环单元(Gated Recurrent Unit,GRU)相结合,提出了一种基于eLU激活函数的CNN-GRU网络模型,用于轴承故障诊断。用凯斯西储大学轴承故障数据集对该网络进行验证,预测准确率达到99.93%,相较于其他算法更具优越性。
Abstract:
Traditional convolutional neural network adopts sigmoid function or tanh function as activation function,which cause gradient disappears and over-fitting. And ReLU function often leads to mean shift. In rolling bearing fault diagnosis, fault data is usually time series signals. In order to solve the problem about activation function and improve the ability of extract the timing features , a CNN-GRU (Gated Recurrent Unit,GRU) model based on ELU activation function is proposed for bearing fault diagnosis by combining CNN with GRU. The network is verified by the bearing fault data set of Case Western Reserve University. The prediction accuracy reaches 99.93%, which is superior to other algorithms.

参考文献/References:

[1]李少波,杨静,王铮,等.缺陷检测技术的发展与应用研究综述[J].自动化学报,2020,46(11):2319-2336.

[2]赵宇凯,徐高威,刘敏.基于VGG16迁移学习的轴承故障诊断方法[J].航天器环境工程,2020,37(05):446-451.
[3]许子非,缪维跑,李春,等.基于变分模态分解与卷积神经网络的轴承故障诊断[J].热能动力工程,2021,36(03):55-63.
[4]陈志强,陈旭东,José Valente de Olivira,等.深度学习在设备故障预测与健康管理中的应用[J].仪器仪表学报,2019,40(09):206-226.
[5]郭一帆,唐家银.基于机器学习算法的寿命预测与故障诊断技术的发展综述[J].计算机测量与控制,2019,27(03):7-13.
[6] 张西宁,余迪,刘书语.基于迁移学习的小样本轴承故障诊断方法研究[J/OL].西安交通大学学报,2021(10):1-8.
[7]徐先峰,王研,刘阿慧,等.基于CNN并引入EMD预处理机制的滚动轴承故障诊断方法[J].工业仪表与自动化装置,2020(04):7-11.
[8]庞俊,刘鑫,段敏霞,等.基于改进卷积神经网络轴承故障诊断[J].组合机床与自动化加工技术,2021(03):66-69.
[9]王崇宇,郑召利,刘天源,等.基于卷积神经网络的汽轮机转子不平衡与不对中故障检测方法研究[J].中国电机工程学报,2021,41(07):2417-2427.
[10] 薛阳,雷文平,岳帅旭,等.多模态学习方法在滚动轴承故障诊断中的应用[J/OL].机械科学与技术:1-6[2021-07-06].
[11]王昊,肖慧灵,王丽亚,等.一种基于改进迁移策略与膨胀卷积神经网络的轴承故障诊断方法[J/OL].工业工程与管理:1-11[2021-07-06].
[12]张龙,徐天鹏,王朝兵,等.基于卷积门控循环网络的齿轮箱故障诊断[J/OL].吉林大学学报(工学版):1-10[2021-07-06].
[13]郭玲,徐青山,郑乐.基于TCN-GRU模型的短期负荷预测方法[J].电力工程技术,2021,40(03):66-71.
[14]陈仁祥,黄鑫,杨黎霞,等.基于卷积神经网络和离散小波变换的滚动轴承故障诊断[J].振动工程学报,2018,31(05):883-891.

相似文献/References:

[1]尹新权,王 珺,张亚萍.基于模糊理论的柴油机故障诊断专家系统[J].工业仪表与自动化装置,2015,(01):111.
 YIN Xinquan,WANG Jun,ZHANG Yaping.Fault diagnostic expert system of diesel engine based on fuzzy theory[J].Industrial Instrumentation & Automation,2015,(06):111.
[2]孟文俊a,徐光华a,b,等.基于LabVIEW的滚动轴承非平稳过程监测诊断及性能评估系统的开发[J].工业仪表与自动化装置,2015,(02):18.
 MENG Wenjuna,XU Guanghuaa,b,et al.Development of non-stationary process for rolling bearing fault diagnosis and performance evaluation system based on LabVIEW[J].Industrial Instrumentation & Automation,2015,(06):18.
[3]李 茜,王延年.基于普通铣床数控化的S7-300 PLC远程监控和故障诊断系统设计[J].工业仪表与自动化装置,2015,(02):49.
 LI Qian,WANG Yannian.Design of remote monitoring and fault diagnosis systembased on the ordinary milling machine of numerical control of S7-300 PLC[J].Industrial Instrumentation & Automation,2015,(06):49.
[4]巴寅亮,王书提,谢 鑫.基于改进的BP神经网络的柴油发动机故障诊断[J].工业仪表与自动化装置,2015,(03):94.
 BA Yinliang,WANG Shuti,XIE Xin.Research of diesel engine fault based on improved BP neural network[J].Industrial Instrumentation & Automation,2015,(06):94.
[5]王江荣,文 晖,黄建华.基于差分进化算法的二次回归在矿井通风机故障诊断中的应用[J].工业仪表与自动化装置,2015,(01):50.
 WANG Jiangrong,WEN Hui,HUANG Jianhua.The two regression in ventilator fault diagnosis application based on difference evolutionary algorithm[J].Industrial Instrumentation & Automation,2015,(06):50.
[6]张卫峰,惠俊军.智能故障诊断技术的现状及展望[J].工业仪表与自动化装置,2017,(05):21.
 ZHANG Weifeng,HUI Junjun.The present situation and prospects of intelligence fault diagnosis technology[J].Industrial Instrumentation & Automation,2017,(06):21.
[7]宫玮丽,梁 波,王晓兰.基于小波包和Hilbert包络分析的隧道掘进机主轴承故障诊断方法研究[J].工业仪表与自动化装置,2018,(02):15.[doi:1000-0682(2018)02-0000-00]
 GONG Weili,LIANG Bo,WANG Xiaolan.Research on fault diagnosis method of main bearing of tunnel boring machine based on wavelet packet and Hilbert envelope analysis[J].Industrial Instrumentation & Automation,2018,(06):15.[doi:1000-0682(2018)02-0000-00]
[8]席 维,白 璘,武奇生.基于经验小波变换和峭度值的滚动轴承故障检测方法[J].工业仪表与自动化装置,2018,(06):26.[doi:1000-0682(2018)06-0000-00]
 XI Wei,BAI Lin,WU Qisheng.A novel rolling bearing fault detection method based on empirical wavelet transform and kurtosis value[J].Industrial Instrumentation & Automation,2018,(06):26.[doi:1000-0682(2018)06-0000-00]
[9]张远绪,程换新.基于改进的RBF神经网络的滚动轴承故障诊断[J].工业仪表与自动化装置,2018,(06):31.[doi:1000-0682(2018)06-0000-00]
 ZHANG Yuanxu,CHENG Huanxin.Fault diagnosis of rolling bearing based on improved RBF neural network[J].Industrial Instrumentation & Automation,2018,(06):31.[doi:1000-0682(2018)06-0000-00]
[10]李小斌,王亚云 陈渭红.电动机轴承温度远程在线监测系统设计[J].工业仪表与自动化装置,2018,(06):62.[doi:1000-0682(2018)06-0000-00]
 LI Xiaobin,WANG Yayun,CHEN Weihong.Design of a remote on-line detection system for motor bearings temperature[J].Industrial Instrumentation & Automation,2018,(06):62.[doi:1000-0682(2018)06-0000-00]
[11]张海龙,袁德成.基于多尺度CNN-BiLSTM的轴承故障诊断[J].工业仪表与自动化装置,2022,(03):75.[doi:10.19950/j.cnki.cn61-1121/th.2022.03.016]
 ZHANG Hailong,YUAN Decheng.Bearing fault diagnosis based on multi-scale CNN-BiLSTM[J].Industrial Instrumentation & Automation,2022,(06):75.[doi:10.19950/j.cnki.cn61-1121/th.2022.03.016]
[12]李 娜,曹丽明.一种风力发电机轴承故障智能诊断方法[J].工业仪表与自动化装置,2022,(05):103.[doi:10.19950/j.cnki.cn61-1121/th.2022.05.019]
 LI Na,CAO Liming.An intelligent diagnosis method for wind turbine bearing fault[J].Industrial Instrumentation & Automation,2022,(06):103.[doi:10.19950/j.cnki.cn61-1121/th.2022.05.019]

备注/Memo

备注/Memo:
收稿日期:2021-07-07
作者简介:张伟(1975),男,重庆开县人,博士,副教授,研究方向为工业智能控制。
更新日期/Last Update: 1900-01-01