基于超球优化支持向量数据描述的滚动轴承故障检测
|
摘要
在仅有轴承正常运行数据的小样本情况下,支持向量数据描述(SVDD)能通过对多维特征的融合实现滚动轴承的故障检测与状态评估,但特征向量空间分布的复杂程度会直接影响SVDD的效果。为此,提出了一种基于超球优化支持向量数据描述的滚动轴承故障检测方法,通过超球优化改善特征向量的空间分布以降低数据描述任务的难度,进而使得超球优化SVDD能更有效地识别出滚动轴承故障。多组试验表明:在不同转速、不同测点、不同类型的滚动轴承故障下,超球优化SVDD比传统的SVDD方法效果更优。
In the case of small sample size problems where only the operating data of healthy rolling bearings are available, the support vector data description(SVDD) method was applied to the rolling bearings fault detection and condition evaluation commendably by fusing multidimensional features. However, the complexity of the feature vector space distribution will directly affects the results of SVDD. Aiming at this, a novel rolling bearing fault detection method called hyper-sphere optimization support vector data description(hoSVDD) was proposed. The spatial distribution of feature vectors was improved by the hyper-sphere optimization so that the difficulty in data description was reduced. Hence, the hoSVDD is more suitable for rolling bearing fault detection. Multi-group rolling bearing tests show that: under the conditions of different speeds, different test points, and different types of rolling bearings faults, the proposed hoSVDD performs better than the traditional SVDD method.
In the case of small sample size problems where only the operating data of healthy rolling bearings are available, the support vector data description(SVDD) method was applied to the rolling bearings fault detection and condition evaluation commendably by fusing multidimensional features. However, the complexity of the feature vector space distribution will directly affects the results of SVDD. Aiming at this, a novel rolling bearing fault detection method called hyper-sphere optimization support vector data description(hoSVDD) was proposed. The spatial distribution of feature vectors was improved by the hyper-sphere optimization so that the difficulty in data description was reduced. Hence, the hoSVDD is more suitable for rolling bearing fault detection. Multi-group rolling bearing tests show that: under the conditions of different speeds, different test points, and different types of rolling bearings faults, the proposed hoSVDD performs better than the traditional SVDD method.
引文
[1] YU J.Bearing performance degradation assessment using locality preserving projections and Gaussian mixture models[J].Mechanical Systems and Signal Processing,2011,25(7):2573-2588.
[2] 程军圣,史美丽,杨宇.基于LMD与神经网络的滚动轴承故障诊断方法[J].振动与冲击,2010,29(8):141-144.CHENG Junsheng,SHI Meili,YANG Yu.Roller bearing fault diagnosis method based on LMD and neural network[J].Journal of Vibration and Shock,2010,29(8):141-144.
[3] 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,et al.Fault pattern recognition technique for roller bearing acoustic emission based on harmonic wavelet packet and BP neural network[J].Journal of Vibration and Shock,2010,29(10):162-165.
[4] 郑近德,程军圣,杨宇.基于改进的ITD和模糊熵的滚动轴承故障诊断方法[J].中国机械工程,2012,23(19):2372-2377.ZHENG Jinde,CHENG Junsheng,YANG Yu.A Rolling bearing fault diagnosis method based on improved ITD and fuzzy entropy[J].China Mechanical Engineering,2012,23(19):2372-2377.
[5] 赵志宏,杨绍普.一种基于样本熵的轴承故障诊断方法[J].振动与冲击,2012,31(6):136-140.ZHAO Zhihong,YANG Shaopu.Sample entropy-based roller bearing fault diagnosis method[J].Journal of Vibration and Shock,2012,31(6):136-140.
[6] 朱启兵,杨宝,黄敏.基于核映射稀疏表示分类的轴承故障诊断[J].振动与冲击,2013,32(11):30-34.ZHU Qibing,YANG Bao,HUANG Min.Bearing fault diagnosis using kernel-mapping sparse representation classification algorithm[J].Journal of Vibration and Shock,2013,32(11):30-34.
[7] CHEN S L,CRAIG M,WOOD R J K,et al.Bearing condition monitoring using multiple sensors and integrated data fusion techniques[C]// Proceedings of the Ninth International Conference in Vibrations in Rotating Machinery.Oxford:[s.n.],2008.
[8] 李巍华,戴炳雄,张绍辉.基于小波包熵和高斯混合模型的轴承性能退化评估[J].振动与冲击,2013,32(21):35-40.LI Weihua,DAI Bingxiong,ZHANG Shaohui.Bearing performance degradation assessment based on Wavelet packet entropy and Gaussian mixture model[J].Journal of Vibration and Shock,2013,32(21):35-40.
[9] HUANG R,XI L,LI X,et al.Residual life predictions for ball bearings based on self-organizing map and back propagation neural network methods[J].Mechanical Systems & Signal Processing,2007,21(1):193-207.
[10] 张全德,陈果,林桐,等.基于自组织神经网络的滚动轴承状态评估方法[J].中国机械工程,2017,28(5):550-558.ZHANG Quande,CHEN Guo,LIN Tong,et al.Condition assessment for rolling bearings based on SOM[J].China Mechanical Engineering,2017,28(5):550-558.
[11] TAX D M J,DUIN R P W.Support vector domain description[J].Pattern Recognition Letters,1999,20(11):1191-1199.
[12] 潘玉娜,陈进.小波包-支持向量数据描述在轴承性能退化评估中的应用研究[J].振动与冲击,2009,28(4):164-167.PAN Yuna,CHEN Jin.Wavelet package-support vector data description applied in bearing performance degradation assessment[J].Journal of Vibration and Shock,2009,28(4):164-167.
[13] SHEN Z J,HE Z J,CHEN X F,et al.A monotonic degradation assessment index of rolling bearings using fuzzy support vector data description and running time[J].Sensors,2012,12(8):10109-10135.
[14] 郝腾飞,陈果.基于小球大间隔方法的机械故障检测[J].中国机械工程,2012,23(15):1765-1770.HAO Tengfei,CHEN Guo.Machinery fault detection based on a small sphere and large margin approach[J].China Mechanical Engineering,2012,23(15):1765-1770.
[15] 吴定海,张培林,王怀光,等.基于多核支持向量数据描述的单类分类方法[J].计算机工程,2013,39(5):165-168.WU Dinghai,ZHANG Peilin,WANG Huaiguang,et al.One-class classification method based on multi-kernel support vector data description[J].Computer Engineering,2013,39(5):165-168.
[16] 卢明,刘黎辉,吴亮红.多核支持向量数据描述分类方法研究[J].计算机工程与应用,2016,52(18):68-73.LU Ming,LIU Lihui,WU Lianghong.Research on multi-kernel support vector data description method of classification[J].Computer Engineering and Applications,2016,52(18):68-73.
[17] CORTES C,VAPNIK V.Support-vector networks[J].Machine Learning,1995,20(3):273-297.
[18] 周志华.机器学习[M].北京:清华大学出版社,2016.
[19] CHEN G,HAO T F,WANG H F,et al.Sensitivity analysis and experimental research on ball bearing early fault diagnosis based on testing signal from casing[J].Journal of Dynamic Systems,Measurement,and Control,2014,136(6):061009.
[2] 程军圣,史美丽,杨宇.基于LMD与神经网络的滚动轴承故障诊断方法[J].振动与冲击,2010,29(8):141-144.CHENG Junsheng,SHI Meili,YANG Yu.Roller bearing fault diagnosis method based on LMD and neural network[J].Journal of Vibration and Shock,2010,29(8):141-144.
[3] 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,et al.Fault pattern recognition technique for roller bearing acoustic emission based on harmonic wavelet packet and BP neural network[J].Journal of Vibration and Shock,2010,29(10):162-165.
[4] 郑近德,程军圣,杨宇.基于改进的ITD和模糊熵的滚动轴承故障诊断方法[J].中国机械工程,2012,23(19):2372-2377.ZHENG Jinde,CHENG Junsheng,YANG Yu.A Rolling bearing fault diagnosis method based on improved ITD and fuzzy entropy[J].China Mechanical Engineering,2012,23(19):2372-2377.
[5] 赵志宏,杨绍普.一种基于样本熵的轴承故障诊断方法[J].振动与冲击,2012,31(6):136-140.ZHAO Zhihong,YANG Shaopu.Sample entropy-based roller bearing fault diagnosis method[J].Journal of Vibration and Shock,2012,31(6):136-140.
[6] 朱启兵,杨宝,黄敏.基于核映射稀疏表示分类的轴承故障诊断[J].振动与冲击,2013,32(11):30-34.ZHU Qibing,YANG Bao,HUANG Min.Bearing fault diagnosis using kernel-mapping sparse representation classification algorithm[J].Journal of Vibration and Shock,2013,32(11):30-34.
[7] CHEN S L,CRAIG M,WOOD R J K,et al.Bearing condition monitoring using multiple sensors and integrated data fusion techniques[C]// Proceedings of the Ninth International Conference in Vibrations in Rotating Machinery.Oxford:[s.n.],2008.
[8] 李巍华,戴炳雄,张绍辉.基于小波包熵和高斯混合模型的轴承性能退化评估[J].振动与冲击,2013,32(21):35-40.LI Weihua,DAI Bingxiong,ZHANG Shaohui.Bearing performance degradation assessment based on Wavelet packet entropy and Gaussian mixture model[J].Journal of Vibration and Shock,2013,32(21):35-40.
[9] HUANG R,XI L,LI X,et al.Residual life predictions for ball bearings based on self-organizing map and back propagation neural network methods[J].Mechanical Systems & Signal Processing,2007,21(1):193-207.
[10] 张全德,陈果,林桐,等.基于自组织神经网络的滚动轴承状态评估方法[J].中国机械工程,2017,28(5):550-558.ZHANG Quande,CHEN Guo,LIN Tong,et al.Condition assessment for rolling bearings based on SOM[J].China Mechanical Engineering,2017,28(5):550-558.
[11] TAX D M J,DUIN R P W.Support vector domain description[J].Pattern Recognition Letters,1999,20(11):1191-1199.
[12] 潘玉娜,陈进.小波包-支持向量数据描述在轴承性能退化评估中的应用研究[J].振动与冲击,2009,28(4):164-167.PAN Yuna,CHEN Jin.Wavelet package-support vector data description applied in bearing performance degradation assessment[J].Journal of Vibration and Shock,2009,28(4):164-167.
[13] SHEN Z J,HE Z J,CHEN X F,et al.A monotonic degradation assessment index of rolling bearings using fuzzy support vector data description and running time[J].Sensors,2012,12(8):10109-10135.
[14] 郝腾飞,陈果.基于小球大间隔方法的机械故障检测[J].中国机械工程,2012,23(15):1765-1770.HAO Tengfei,CHEN Guo.Machinery fault detection based on a small sphere and large margin approach[J].China Mechanical Engineering,2012,23(15):1765-1770.
[15] 吴定海,张培林,王怀光,等.基于多核支持向量数据描述的单类分类方法[J].计算机工程,2013,39(5):165-168.WU Dinghai,ZHANG Peilin,WANG Huaiguang,et al.One-class classification method based on multi-kernel support vector data description[J].Computer Engineering,2013,39(5):165-168.
[16] 卢明,刘黎辉,吴亮红.多核支持向量数据描述分类方法研究[J].计算机工程与应用,2016,52(18):68-73.LU Ming,LIU Lihui,WU Lianghong.Research on multi-kernel support vector data description method of classification[J].Computer Engineering and Applications,2016,52(18):68-73.
[17] CORTES C,VAPNIK V.Support-vector networks[J].Machine Learning,1995,20(3):273-297.
[18] 周志华.机器学习[M].北京:清华大学出版社,2016.
[19] CHEN G,HAO T F,WANG H F,et al.Sensitivity analysis and experimental research on ball bearing early fault diagnosis based on testing signal from casing[J].Journal of Dynamic Systems,Measurement,and Control,2014,136(6):061009.