基于多尺度卷积神经网络的滚动轴承故障诊断方法

摘要
图/表
参考文献(27)
相关文章 (15)

全文: PDF (4167 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要为提升滚动轴承在大噪声、变载荷及复杂工况下故障诊断的准确率，考虑被采信号具有时间多尺度特性，提出多尺度卷积神经网络（MTSC-CNN），开发一种“端到端”的故障诊断系统。为验证MTSC-CNN方法的有效性，通过实验数据，对11种含故障类型、损伤程度不同以及4种存在故障混合的轴承状态进行识别。结果表明：考虑单一时间尺度提取时，因信息缺失导致模型性能欠佳；过多的时间尺度将产生信息过提取，继而增加模型复杂度，且弱化模型诊断能力。与现有方法相比，MTSC-CNN模型在复杂环境下性能更佳。此外，基于可视化技术，表明由于不同尺度所学习特征存在互补性，而使模型具有较强的鲁棒性。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	许子非
	金江涛
	李春

关键词 ：故障诊断, 卷积神经网络(CNN), 轴承, 多尺度

Abstract：In order to improving the accuracy of rolling bearings fault diagnosis in complex working environments, and considering the multiple time scale characteristic of the measured signals, a new algorithm named multiple time scale characteristic extracted convolutional neural network, (MTSC-CNN) was proposed to develop an end-to-end fault diagnosis system.The proposed MTSC-CNN was used to realize fault identification under 11 working conditions of the tested bearings, including different fault types and damage degrees for verifying the effectiveness of the proposed model.The results show that when only single time scale is considered, the performance of the model is poor due to lack of information.Too larger time scale will lead to over-extraction of information, which will increase the computational time and weaken the diagnostic capability of the model.Compared with the existing methods, the MTSC-CNN model has better performance under variable load and noise conditions.In addition, the results of neural network visualization also show that the features learned at different scales are complementary to improve the robustness of the model.

Key words： fault diagnosis convolutional neural network(CNN) bearing multiscale

收稿日期: 2020-04-23 出版日期: 2021-09-28

引用本文:

许子非，金江涛，李春. 基于多尺度卷积神经网络的滚动轴承故障诊断方法[J]. 振动与冲击, 2021, 40(18): 212-220.
XU Zifei,JIN Jiangtao,LI Chun. New method for the fault diagnosis of rolling bearings based on a multiscale convolutional neural network. JOURNAL OF VIBRATION AND SHOCK, 2021, 40(18): 212-220.

链接本文:

http://jvs.sjtu.edu.cn/CN/ 或 http://jvs.sjtu.edu.cn/CN/Y2021/V40/I18/212

［1］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.
ZHUANG Yuxuan，LI Qi，YANG Bingru， et al.An end-to-end approach for bearing fault diagnosis based on LSTM ［J］.Noise and Vibration Control, 2019，39(6): 187-193.

［2］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．
QU Jianling, YU Lu, YUAN Tao, et al.Adaptive fault diagnosis algorithm for rolling bearings based on the one-dimensional neural network ［J］.Chinese Journal of Scientific Instrument, 2018, 39(7):134-143.

［3］贺岩松，黄毅，徐中明，等．基于小波奇异熵与SOFM 神经网络的电机轴承故障识别［J］．振动与冲击，2017，36
(10)：217-223．
HE Yansong，HUANG Yi，XU Zhongming，et al． Motor bearing fault identification based on the wavelet singular entropy and SOFM neural network［J］． Journal of Vibration and Shock，2017，36(10)：217-223．

［4］SONG L, ZHANG X R, SU L W, et al.Fault diagnosis approach for incipient bearing fault in wind turbine under variable conditions［J］.Applied Mechanics & Materials, 2014, 599/600/601:312-320.

［5］LIU W Y, GAO Q W, YE G, et al.A novel wind turbine bearing fault diagnosis method based on Integral Extension LMD［J］.Measurement, 2015, 74:70-77.

［6］ZHANG M, JIANG Z N, FENG K.Research on variational mode decomposition in rolling bearings fault diagnosis of the multistage centrifugal pump［J］.Mechanical Systems and Signal Processing, 2017, 93:460-493.

［7］CHEN X J, YANG Y M, CUI Z X, et al.Vibration fault diagnosis of wind turbine based on variational mode decomposition and energy entropy［J］.Energy, 2019，174：1100-1109.

［8］LI J M, YAO X F, WANG X D, et al.Multiscale local feature learning based on BP network for rolling bearing intelligent fault diagnosis［J］.Measurement, 2020, 153: 107419.

［9］BENGIO Y, COURVILLE A, VINCENT P.Representation learning: a review and new perspectives ［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(8):1798-1828.

［10］GUO Y, WU Z Y, JI Y.A hybrid deep representation learning model for time series classification and prediction［C］∥ 2017 3rd International Conference on Big Data Computing and Communications (BIGCOM).Chengdu: IEEE, 2017.

［11］肖熊，王健翔，张勇军，等.一种用于轴承故障诊断的二维卷积神经网络优化方法［J］.中国电机工程学报，2019，39（15）：4558-4567.
XIAO Xiong, WANG Jianxiang, ZHANG Yongjun, et al.A two dimensional convolutional neural network optimization method for bearing fault diagnosis ［J］.Proceedings of CSEE, 2019,39(15):4458-4567.

［12］CHEN Z Q, LI C, REN-VINICIO S.Gearbox fault identification and classification with convolutional neural networks［J］.Shock and Vibration, 2015, 2015(5):390134.

［13］WANG J J, ZHUANG J F, DUAN L X, et al.A multi-scale convolution neural network for featureless fault diagnosis［C］∥ 2016 International Symposium on Flexible Automation (ISFA).Cleveland： IEEE, 2016.

［14］CHEN Z Y, GRYLLIAS K, LI W H.Mechanical fault diagnosis using convolutional neural networks and extreme learning machine ［J］.Mechanical Systems and Signal Processing, 2019,133：106272.

［15］LI J M, YAO X F, WANG X D, et al.Multiscale local features learning based on BP neural network for rolling bearing intelligent fault diagnosis［J］.Measurement, 2020,153:107419.

［16］CAI J, GACHOT C.Feature extraction of rolling bearing fault signal based on local mean decomposition and Teager energy operator［J］.Industrial Lubrication and Tribology, 2017, 69(6): 872-880.

［17］INCE T, KIRANYAZ S , EREN L , et al.Real-time motor fault detection by 1D convolutional neural networks［J］.IEEE Transactions on Industrial Electronics, 2016, 63(11):7067-7075.

［18］ZHANG W, LI C H, PENG G L, et al.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.

［19］ZHANG L, XIONG G L, LIU H S, et al.Bearing fault diagnosis using multi-scale entropy and adaptive neuro-fuzzy inference［J］.Expert Systems with Applications, 2010, 37(8): 6077-6085.

［20］HUANG W Y, CHENG J S, YANG Y, et al.An improved deep convolutional neural network with multi-scale information for bearing fault diagnosis ［J］.Neurocomputing, 2019, 359:77-92.

［21］LIU H H, HAN M H.A fault diagnosis method based on local mean decomposition and multi-scale entropy for roller bearings［J］.Mechanism and Machine Theory, 2014, 75:67-78.

［22］LUO Y, CHENG Y, UZUNER O, et al.Segment convolutional neural networks (Seg-CNNs) for classifying relations in clinical notes［J］.Journal of the American Medical Informatics Association, 2017, 25(1):93.

［23］KUO C C J.Understanding convolutional neural networks with a mathematical model ［J］.Journal of Visual Communication & Image Representation, 2016, 41: 406-413.

［24］Case Western Reserve University.Bearing data center (seeded fault test data)［EB/OL］.(2013-08-16)［2017-01-02］.http:∥csegroups.case.cedu/bearingdatacenter/home.

［25］WANG B, LEI Y G, LI N P, et al.A hybrid prognostics approach for estimating remaining useful life of rolling element bearings［J］.IEEE Transactions on Reliability, 2018,69(1): 401-412.

［26］CHEN X J, YANG Y M, CUI Z X, et al.Vibration fault diagnosis of wind turbines based on variational mode decomposition and energy entropy［J］.Energy, 2019, 174:1100-1109.

［27］LIANG P F, DENG C, WU J, et al.Compound fault diagnosis of gearboxes via multi-label convolutional neural network and wavelet transform［J］.Computers in Industry, 2019, 113:103-132.