全矢ITD和KPCA结合的滚动轴承故障诊断
|
摘要
针对在滚动轴承故障检测和诊断中获取的单通道信息不全面、不准确等问题,提出了全矢本征时间尺度分解(ITD)和核主元分析(KPCA)相结合的方法以进行故障检测与诊断。首先采用全矢ITD对正常运行状态下的同源双通道原始样本数据进行信息融合,得到全矢融合后的主振矢数据,并建立KPCA模型,克服了单通道振动信号信息不完整的缺点。然后运用KPCA模型对待测样本数据进行在线监控,当该模型的T2和SPE统计量超过已设定的控制限时,采用全矢Hilbert包络分析提取故障数据的特征频率以进行故障诊断。实验结果表明,该方法既能较好地检测出滚动轴承的运行状态,又能准确有效地诊断故障类型。
It proposes a method which combines full vector intrinsic time scale decomposition(ITD)and the kernel principal component analysis(KPCA)for fault detection and diagnosis in view of the problem of incomplete and inaccurate information about single channel information obtained in the detection and diagnosis of rolling bearings. This method firstly uses the full vector ITD to fuse the homologous dual-channel original sample data in normal operation state,obtains the main vibration vector data after full vector fusion to overcome the disadvantages of incomplete single channel vibration signal,and establishes KPCA model. Then,the sample data is monitored online using the KPCA model,when the model's T2 and SPE statistics exceed the set limit,full vector Hilbert envelope demodulation is carried out to extract the feature frequency of fault data for the fault diagnosis. The experimental results show that the method can not only detect the running state of rolling bearing,but also can identify the fault type accurately and effectively.
It proposes a method which combines full vector intrinsic time scale decomposition(ITD)and the kernel principal component analysis(KPCA)for fault detection and diagnosis in view of the problem of incomplete and inaccurate information about single channel information obtained in the detection and diagnosis of rolling bearings. This method firstly uses the full vector ITD to fuse the homologous dual-channel original sample data in normal operation state,obtains the main vibration vector data after full vector fusion to overcome the disadvantages of incomplete single channel vibration signal,and establishes KPCA model. Then,the sample data is monitored online using the KPCA model,when the model's T2 and SPE statistics exceed the set limit,full vector Hilbert envelope demodulation is carried out to extract the feature frequency of fault data for the fault diagnosis. The experimental results show that the method can not only detect the running state of rolling bearing,but also can identify the fault type accurately and effectively.
引文
[1]李志农,刘卫兵,肖尧先.基于局域均值分解包络谱和SVM的滚动轴承故障诊断方法研究[J].机械设计与制造,2011(11):170-172.(Li Zhi-nong,Liu Wei-bing,Xiao Yao-xian.Fault diagnosis method ofrolling bearing based on local mean decomposition envelope spectrumand SVM[J].Machinery Design&Manufacture,2011(11):170-172.)
[2]Bernhard Sch lkopf,Alexander Smola,Klaus-Robert Müller.Nonlinear component analysis as a kernel eigenvalue problem[J].Neural Computation,1998,10(5):1299-1319.
[3]Wang Ai-juan,Zhang Duan-jin,Jie Xiao-jing.Fault detection based on kernel principal component analysis[J].Engineering Structures,2013,32(11):3683-3691.
[4]Li Ping,Li Xue-jun,Jiang Ling-li.Fault diagnosis for motor rotor based on KPCA-SVM[J].Applied Mechanics&Materials,2011,143-144:680-684.
[5]王华忠,俞金寿.基于核函数PCA的非线性过程实时监控方法[J].华东理工大学学报:自然科学版,2005(6):783-786+824.(Wang Hua-zhong,Yu Jin-shou.On-line monitoring of nonlinear processes based on kernel principal component analysis[J].Journal of East China University of Science and Technology:Natural Science Edition,2005(6):783-786+824.)
[6]赵忠盖,刘飞.基于稀疏核主元分析的在线非线性过程监控[J].化工学报,2008(7):1773-1777.(Zhao Zhong-gai,Liu Fei.On-line nonlinear process monitoring based on sparse kernel principal component analysis[J].Journal of Chemical Industry,2008(7):1773-1777.)
[7]Frei M G,Ivan O.Intrinsic time-scale decomposition:time-frequencyenergy a nalysis and real-time filtering of non-stationary signals[J].Proceedings of the Royal Society A:Mathematical,Physical and Engineering Science,2007(463):321-342.
[8]王洪明.基于ITD-矢Teager能量算子的滚动轴承故障诊断方法研究[D].郑州:郑州大学,2016.(Wang Hong-ming.Research on fault diagnosis method of rolling bearing based on ITD and vector teager energy[D].Zhengzhou:Zhengzhou University,2016.)
[9]巩晓赟,韩捷,陈宏.全矢小波包-包络分析方法及其在齿轮故障诊断中的应用[J].振动与冲击,2012,31(12):92-95.(Gong Xiao-yun,Han Jie,Chen Hong.Application of full vector wavelet packet and envelope analysis method in gear fault diagnosis[J].Journal of Vibration and Shock,2012,31(12):92-95.)
[10]韩捷,石来德.全矢谱技术及工程应用[M].北京:机械工业出版社.2008(9).(Han Jie,Shi Lai-de.Technology of Full Vector Spectrum and Engineering Application[M].Beijing:China Machine Press,2008(9).)
[11]罗颂荣,程军圣,杨宇.基于本征时间尺度分解和变量预测模型模式识别的机械故障诊断[J].振动与冲击,2013(13):43-48.(Luo Song-rong,Cheng Jun-shen,Yang Yu.Machine fault diagnosis method using ITD and variable predictive model-based class discrimination[J].Journal of and Shock,2015(5):806-811.)
[2]Bernhard Sch lkopf,Alexander Smola,Klaus-Robert Müller.Nonlinear component analysis as a kernel eigenvalue problem[J].Neural Computation,1998,10(5):1299-1319.
[3]Wang Ai-juan,Zhang Duan-jin,Jie Xiao-jing.Fault detection based on kernel principal component analysis[J].Engineering Structures,2013,32(11):3683-3691.
[4]Li Ping,Li Xue-jun,Jiang Ling-li.Fault diagnosis for motor rotor based on KPCA-SVM[J].Applied Mechanics&Materials,2011,143-144:680-684.
[5]王华忠,俞金寿.基于核函数PCA的非线性过程实时监控方法[J].华东理工大学学报:自然科学版,2005(6):783-786+824.(Wang Hua-zhong,Yu Jin-shou.On-line monitoring of nonlinear processes based on kernel principal component analysis[J].Journal of East China University of Science and Technology:Natural Science Edition,2005(6):783-786+824.)
[6]赵忠盖,刘飞.基于稀疏核主元分析的在线非线性过程监控[J].化工学报,2008(7):1773-1777.(Zhao Zhong-gai,Liu Fei.On-line nonlinear process monitoring based on sparse kernel principal component analysis[J].Journal of Chemical Industry,2008(7):1773-1777.)
[7]Frei M G,Ivan O.Intrinsic time-scale decomposition:time-frequencyenergy a nalysis and real-time filtering of non-stationary signals[J].Proceedings of the Royal Society A:Mathematical,Physical and Engineering Science,2007(463):321-342.
[8]王洪明.基于ITD-矢Teager能量算子的滚动轴承故障诊断方法研究[D].郑州:郑州大学,2016.(Wang Hong-ming.Research on fault diagnosis method of rolling bearing based on ITD and vector teager energy[D].Zhengzhou:Zhengzhou University,2016.)
[9]巩晓赟,韩捷,陈宏.全矢小波包-包络分析方法及其在齿轮故障诊断中的应用[J].振动与冲击,2012,31(12):92-95.(Gong Xiao-yun,Han Jie,Chen Hong.Application of full vector wavelet packet and envelope analysis method in gear fault diagnosis[J].Journal of Vibration and Shock,2012,31(12):92-95.)
[10]韩捷,石来德.全矢谱技术及工程应用[M].北京:机械工业出版社.2008(9).(Han Jie,Shi Lai-de.Technology of Full Vector Spectrum and Engineering Application[M].Beijing:China Machine Press,2008(9).)
[11]罗颂荣,程军圣,杨宇.基于本征时间尺度分解和变量预测模型模式识别的机械故障诊断[J].振动与冲击,2013(13):43-48.(Luo Song-rong,Cheng Jun-shen,Yang Yu.Machine fault diagnosis method using ITD and variable predictive model-based class discrimination[J].Journal of and Shock,2015(5):806-811.)