|
|
|
| A bearing fault intelligent diagnosis method based on deep convolution neural network and WPT-PWVD |
| HUANG Xin1， CHEN Renxiang1,2， YANG Xing1， ZHANG Xia1， HUANG Yu3， YU Tengwei1 |
1.Chongqing Engineering Laboratory for Transportation Engineering Application Robot, Chongqing Jiaotong University, Chongqing400074, China;
2.The State Key Laboratory of Mechanical Transmission, Chongqing University, Chongqing400030, China;
3.School of Automotive and Transportation Engineering, Xihua University, Chengdu610039, China |
|
|
|
|
Abstract Considering problems of low efficiency, complex process, bad generalization and weak self-adaption of artificial feature extraction in bearing fault diagnosis, a new intelligent fault diagnosis method based on deep convolution neural network (DCNN) and the wavelet package transform (WPT) was proposed.Firstly, the bearing fault signal is self-adaptively decomposed into several frequency bands by WPT and effective high-frequency components are extracted and reconstructed.Secondly, the Hilbert algorithm is used to perform envelope demodulation on the reconstructed signal and the PWVD algorithm is performed on the demodulated signal to obtain time-frequency distribution maps which can reveal main fault information.Finally, DCNN is constructed to automatically extract features from the time-frequency distribution maps, then the Softmax multi-classifier is added in feature output layer to perform fine-tune parameters so that feature learning and fault diagnosis is fused.The method is used to diagnose bearing faults under different working conditions, different degrees and different faults.The results prove that the proposed method has high diagnostic accuracy as well as strong generalization.
|
|
Received: 09 November 2018
Published: 15 August 2020
|
|
|
|
[1] 王红军, 万鹏. 基于EEMD和小波包变换的早期故障敏感特征获取[J]. 北京理工大学学报, 2013, 33(9):945-950.
Wang Hongjun, Wan Peng.Earlier fault sensitive feature acquisition based on EEMD and wavelet packet transform[J].Journal of Beijing University of Technology, 2013, 33 (9): 945-950.
[2]Wang L, Liu Z, Miao Q, et al. Time-frequency analysis based on ensemble local mean decomposition and fast kurtogram for rotating machinery fault diagnosis[J]. Mechanical Systems & Signal Processing, 2018, 103:60-75.
[3]Wang T, Liang M, Li J, et al. Rolling element bearing fault diagnosis via fault characteristic order (FCO) analysis[J]. Mechanical Systems & Signal Processing, 2014, 45(1):139-153.
[4]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(5):94-104.
Leia Guo, Jia Feng, Kong Detong, et al. Opportunities and challenges of mechanical intelligent fault diagnosis under large data [J].Journal of Mechanical Engineering, 2018, 54 (5): 94-104.
[5]Lu C, Wang Z Y, Qin W L, et al. Fault diagnosis of rotary machinery components using a stacked denoising autoencoder-based health state identification[J]. Signal Processing, 2017, 130(C):377-388.
[6]李巍华,单外平,曾雪琼. 基于深度信念网络的轴承故障分类识别[J]. 振动工程学报, 2016, 29(2):340-347.
Li Weihua, Shan Weiping, Zeng Xueqiong.Bearing fault classification and identification based on depth belief network[J].Journal of Vibration Engineering, 2016, 29 (2): 340-347.
[7]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.
Rei Yaguo, Jia Feng, Zhou Xin, et al. Large data health monitoring method for mechanical equipment based on depth learning theory [J].Journal of Mechanical Engineering, 2015, 51 (21): 49-56.
[8]周飞燕, 金林鹏, 董军.卷积神经网络研究综述[J]. 计算机学报, 2017, 40(6):1229-1251.
Zhou Feiyan, Jin Linpeng, Dong Jun. A review of convolutional neural networks[J].Journal of Computer Science, 2017, 40 (6): 1229-1251.
[9] Srivastava N, Hinton G, Krizhevsky A, et al. Dropout: a simple way to prevent neural networks from overfitting[J]. Journal of Machine Learning Research, 2014, 15(1):1929-1958.
[10] Behley J, Steinhage V, Cremers A B. Laser-based segment classification using a mixture of bag-of-words[C]// Ieee/rsj International Conference on Intelligent Robots and Systems. IEEE, 2013:4195-4200.
[11] Han J, Zhang D, Wen S, et al. Two-Stage Learning to Predict Human Eye Fixations via SDAEs[J]. IEEE Trans Cybern, 2016, 2(46):487-498.
[12]曹建军, 张培林, 任国全,等. 提升小波包最优基分解算法及在振动信号降噪中的应用[J]. 振动与冲击, 2008, 27(8):114-116.
Cao Jianjun, Zhang Peilin, Ren Guoquan, et al. Lifting wavelet package decomposing algorithm under the best bases and its applications in de-noising for vibration signal [J].Vibration and Impact, 2008, 27 (8): 114-116.
[13] 武和雷, 朱善安, 林瑞仲,等. 基于能量算子解调法的滚动轴承故障诊断[J]. 农业机械学报, 2003, 34(1):118-120.
Wu He Lei,Zhu Shan'an,Lin Ruizhong,et al. Rolling bearing fault diagnosis based on energy operator demodulation[J].Journal of Agricultural Machinery,2003,34(1): 118-120.
[14]牟伟杰, 石林锁, 蔡艳平,等. 基于 KVMD-PWVD 与 LNMF 的内燃机振动谱图像识别诊断方法[J].振动与冲击, 2017, 36(2):45-51.
Mou Weijie, Shi Linsuo, Cai Yanping, et al. Image recognition and diagnosis method for vibration spectrum of internal combustion engine based on KVMD-PWVD and LNMF [J]. Vibration and impact, 2017, 36 (2): 45-51.
[15]康海英, 栾军英, 郑海起,等. 基于小波包变换和伪魏格纳分布的轴承故障诊断[J]. 军械工程学院学报, 2004, 16(5):5-8.
Kang Haiying,Luan Junying,Zheng Haiqi,et al.Bearing fault diagnosis based on wavelet packet transform and pseudo-Wigner distribution[J].Journal of Ordnance Engineering College,2004,16(5): 5-8.
[16]Sokolova M, Japkowicz N, Szpakowicz S. Beyond accuracy, f-score and ROC: a family of discriminant measures for performance evaluation[C]// Australian Joint Conference on Artificial Intelligence: Advances in Artificial Intelligence. Springer-Verlag, 2006:1015-1021.
[17]Jin X, Zhao M, Chow T W S, et al. Motor Bearing Fault Diagnosis Using Trace Ratio Linear Discriminant Analysis[J]. IEEE Transactions on Industrial Electronics, 2013, 61(5):2441-2451.
[18]I.A. AbuMahfouz. "A comparative study of three artificial neural networks for the detection and classification of gear faults." International Journal of General Systems 34.3(2005):261-277.
|
|
|
|
