为了对旋转机械的故障特征进行自适应提取，实现智能故障诊断，提出了一种基于批量归一化的一维卷积神经网络（convolutional neural networks, CNN）模型。由于卷积神经网络通常应用于二维图像或三维视频领域，故通过将卷积核改进为一维卷积核来实现对采集的一维振动数据的直接卷积，并且采用了批归一化层来防止过拟合，采用HZXT-008小型转子实验台采集的数据对该方法进行验证。试验结果表明该方法平均诊断准确率高达98.43%，并且与其他模型相比稳定性更高。该方法实现了大量样本下旋转机械不同故障类型的故障特征自适应提取与故障类型的准确识别。
Abstract
Here, to adaptively extract fault features of rotating machinery and realize intelligent fault diagnosis, a 1D convolutional neural network (CNN) model based on batch normalization, i.e., BN-1DCNN model was proposed. Because CNN is usually used in fields of 2D image or 3D video, the convolution kernel was changed to 1D one to realize direct convolution for the collected 1D vibration data, and the batch normalization layer was used to prevent over-fitting. Finally, the data collected using HZXT-008 small rotor test platform was used to verify the proposed BN-1DCNN model. The test results showed that the average diagnostic accuracy of the proposed BN-1DCNN model can reach 98.43%, this model is more stable than other models; BN-1DCNN model can realize adaptive fault feature extraction and fault type identification of rotating machinery under a large number of samples.
关键词
深度学习 /
卷积神经网络（CNN） /
旋转机械 /
故障诊断
{{custom_keyword}} /
Key words
deep learning /
convolutional neural networks (CNN) /
rotating machinery /
fault diagnosis
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］程国志.基于时频分析法的煤矿主要通风机故障诊断研究［J］.能源与环保, 2018, 40(8):145-149.
CHENG Zhiguo.Study on fault diagnosis of main fan in coal mine based on time frequency analysis［J］.China Energy and Environmental Protection,2018, 40(8):145-149.
［2］冷永刚, 郑安总, 范胜波.SVD分量包络检测方法及其在滚动轴承早期故障诊断中的研究［J］.振动工程学报, 2014, 27(5):794-800.
LENG Yonggang,ZHENG Anzong,FANG Shengbo.SVD component envelope detection method and its application in early fault diagnosis of rolling bearings［J］.Journal of Vibration Engineering,2014, 27(5):794-800.
［3］郑近德, 陈敏均, 程军圣, 等.多尺度模糊熵及其在滚动轴承故障诊断中的应用［J］.振动工程学报, 2014, 27(1):145-151.
ZHENG Jinde,CHEN Minjun,CHENG Junsheng,et al.Multi-scale fuzzy entropy and its application in rolling bearing fault diagnosis［J］.Journal of Vibration Engineering, 2014, 27(1):145-151.
［4］武哲, 杨绍普, 刘永强.基于多元经验模态分解的旋转机械早期故障诊断方法［J］.仪器仪表学报, 2016, 37(2):241-248.
WU Zhe,YANG Shaopu,LIU Yongqiang.Rotating machinery early fault diagnosis method based on multivariate empirical mode decomposition［J］.Chinese Journal of Scientific Instrument,2016, 37(2):241-248.
［5］蒋永华, 程光明, 阚君武, 等.基于优化的滚动轴承故障诊断［J］.仪器仪表学报, 2013, 34(12):2684-2689.
JIANG Yonghua,CHENG Guangming,KAN Junwu,et al.Fault diagnosis of rolling bearings based on optimization［J］.Chinese Journal of Scientific Instrument,2013, 34(12):2684-2689.
［6］PANDYA D H, UPADHYAY S H, HARSHA S P.Fault diagnosis of rolling element bearing with intrinsic mode function of acoustic emission data using APF-KNN［J］.Expert Systems with Application, 2013, 40(10):4137-4145.
［7］HINTON G E, SALAKHUTDINOV R R.Reducing the dimensionality of data with neural networks［J］.Science, 2006, 313:3504-3507.
［8］LECUN Y, BENGIO Y, HINTON G E.Deep learning［J］.Nature, 2015, 521:436-444.
［9］CIRESAN D C, MEIER U, GAMBARDELLA L M, et al.Convolutional neural network committees for handwritten character classification［C］//2011 International Conference on Document Analysis and Recognition. Beijing: ICDAR, 2011.
［10］SZEGEDY C, LIU W, JIA Y Q, et al.Going deeper with convolutions［C］//2015 IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE,2015.
［11］周奇才, 刘星辰, 赵炯, 等.旋转机械一维深度卷积神经网络故障诊断研究［J］.振动与冲击, 2018, 37(23):31-37.
ZHOU Qicai, LIU Xingchen,ZHAO Jiong,et al.Fault diagnosis for rotating machinery based on 1D depth convolutional neural network［J］.Journal of Vibration and Shock,2018, 37(23):31-37.
［12］李恒, 张氢, 秦仙蓉, 等.基于短时傅里叶变换和卷积神经网络的轴承故障诊断方法［J］.振动与冲击, 2018, 37(19):124-131.
LI Heng,ZHANG Qing,QIN Xianrong,et al.Fault diagnosis method for rolling bearings based on short-time Fourier transform and convolution neural network［J］.Journal of Vibration and Shock, 2018, 37(19):124-131.
［13］HUANG T D, FU S, FENG H N, et al.Bearing fault diagnosis based on shallow multi-scale convolutional neural network with attention［J］.Energies, 2019, 12(20):3937.
［14］IOFFE S, SZEGEDY C.Batch normalization: accelerating deep network training by reducing internal covariate shift［C］// International Conference on International Conference on Machine Learning. Miami: IEEE, 2015.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(1857 KB)
