针对滚动轴承振动数据耦合程度高,信号特征提取和识别模型建立困难的问题,本文提出了一种基于深度学习理论的状态监测方法。提取振动信号的时域、频域和时频域特征构成特征向量。通过稀疏自编码非监督学习网络对输入向量进行特征学习,并将单层网络叠加构成深度神经网络。最后采用少量有标签数据对整个深度神经网络进行微调训练,建立轴承状态监测模型。试验结果表明,提出的方法对于轴承状态识别准确率达到90.86%,且性能退化阶段识别率最高,能满足视情维修的工程需求。
Abstract
Vibration signal of rolling element bearing is in a high degree of coupling, so that the features and recognition model are difficult to build. For solving those problems, we proposed a novel bearing condition monitoring model that is based on deep learning. Time domain, frequency domain and time-frequency domain features are extracted. Then those feature vectors are entered into unsupervised auto-encoder to learning high-level features. At the same, the middle layers of auto-encoder network are stacked into multilayered network. Finally, a small number of labeled training samples are used to fine-turning deep learning network. The bearing condition recognition experiment shows that the proposed method gets the state of the art result, and its high accuracy of performance degradation condition is really helpful for condition-based maintenance.
关键词
深度学习 /
非监督学习 /
滚动轴承 /
视情维修
{{custom_keyword}} /
Key words
Deep Learning /
Unsupervised Learning /
Antifriction Bearing /
Condition-based Maintenance
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 高宏力, 李登万, 许明恒. "基于人工智能的丝杠寿命预测技术." 西南交通大学学报 45.5 (2010): 685-691.
High Grace, Lee Teng-Wan, and Xu Mingheng "Screw life prediction based on artificial intelligence technology." Journal of Southwest Jiaotong University, 45.5 (2010): 685-691.
[2] 朱可恒.滚动轴承振动信号特征提取及诊断方法研究[D]. 大连理工大学,2013
Hengke Zhu. Research on Vibration Signal based Rolling Element Bearing Feature Extraction and Fault Diagnosis Method[D]. DaLian University of Technology, 2013
[3] Subrahmanyam M.,Sujatha C. Using neural networks for the diagnosis of localized defects in ball bearings[J]. Tribology International, 1997, 30(10):739-752.
[4] Samanta B., AL-Balushi K. R. Artificial neural network based fault diagnostics of rolling element bearings using time-domain features [J]. Mechanical Systems and Signal Processing, 2003,17(2):317-328.
[5] Sanz J., Perera R., Huerta C. Fault diagnosis of rotating machinery based on auto-associative neural networks and wavelet transforms[J]. Journal of Sound and Vibration, 2007,302(4-5):981-999.
[6] Yu, Kai, Yuanqing Lin, and John Lafferty. Learning image representations from the pixel level via hierarchical sparse coding. Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on. IEEE, 2011.
[7] Dahl, George E., et al. "Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition." Audio, Speech, and Language Processing, IEEE Transactions on 20.1 (2012): 30-42.
[8] Glorot, Xavier, Antoine Bordes, and Yoshua Bengio. "Domain adaptation for large-scale sentiment classification: A deep learning approach." Proceedings of the 28th International Conference on Machine Learning (ICML-11). 2011.
[9] Hinton, Geoffrey E., Simon Osindero, and Yee-Whye Teh. "A fast learning algorithm for deep belief nets." Neural computation 18.7 (2006): 1527-1554.
[10] Verma, Nishchal K., et al. "Intelligent condition based monitoring of rotating machines using sparse auto-encoders." Prognostics and Health Management (PHM), 2013 IEEE Conference on. IEEE, 2013.
[11] Lemme, Andre, René Felix Reinhart, and Jochen Jakob Steil. "Efficient online learning of a non-negative sparse autoencoder." ESANN. 2010.
[12] Fink, Olga, and Ulrich Weidmann. "Using Deep Belief Networks for Predicting Railway Operations Failures." Advances in Risk and Reliability Technology Symposium. 2013.
[13] Moody, John Matali. "Process monitoring with restricted Boltzmann machines." (2014).
[14] 赵元喜, 胥永刚, 高立新, 等. 基于谐波小波包和 BP 神经网络的滚动轴承声发射故障模式识别技术[J]. 振动与冲击, 2010, 29(10): 162-165.
Zhao YuanXi, Xu YongGang, Gao LiXin. Acoustic emission fault recognition of rolling element bearing based on harmonic wavelet package and BP neural network[J]. Journal of Vibration and Shock. 2010, 29(10): 162-165.
[15] 董绍江, 汤宝平, 张焱. 基于非广延小波特征尺度熵和支持向量机的轴承状态识别[J]. 振动与冲击, 2012, 31(15): 50-54.
Hong Jie, Han Lei, Miao Xuewen, Ma Yanhong.
Bearing running state recognition based on non-extensive wavelet feature scale entropy and support vector machine [J]. Journal of Vibration and Shock. 2012, 31(15): 50-54.
[16] Nectoux P, Gouriveau R, Medjaher K, et al. PRONOSTIA: An experimental platform for bearings accelerated degradation tests[C]//IEEE International Conference on Prognostics and Health Management, PHM'12. IEEE Catalog Number: CPF12PHM-CDR, 2012: 1-8.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}