基于流形学习的机械状态识别方法研究
|
摘要
机械装备是一个复杂的非线性系统，由于振动信号包含着描述系统运行状态的丰富信息，振动信号分析是目前常用的机械系统故障诊断方法之一。一般通过在机械构件的关键部位测取振动信号，通过信号处理获得表征机械运行状态的特征指标，常用的信号处理方法有时域分析、频域分析及时频分析。但是实际测得的信号往往存在着非线性、非高斯分布的特点，加上故障形态的多变性，使得传统的信号处理方法在复杂系统故障诊断中存在着一定的局限性，有必要探索用于复杂非线性系统的故障诊断新方法，以防范于未然，减少损失。
流形学习算法是近年来模式识别领域的研究热点之一，其本质在于通过一定的非线性映射将高维空间的数据结构在低维空间中进行表示，同时最大程度的保留高维空间数据的有用信息。因此，可以利用流形学习的优点，用以处理从时域、频域以及时频域中提取的多维信号特征或者处理由多个传感器获取的多源信号，实现机械运行状态的识别。然而，研究中发现存在以下问题：(1)噪声直接影响着流形学习算法的稳健性；(2)参数选择影响了算法的特征提取效果；(3)某些算法对高维数据结构信息保持不完整等。为此，本文从流形学习算法的基础理论出发，研究算法在机械系统状态识别、趋势分析中的噪声敏感性以及参数选择等问题，具体工作如下：
(1)传统时域降噪方法需要消耗大量的计算时间及存储空间，不利于实现机械系统在线诊断。提出直接对特征样本空间进行降噪的方法，理论分析了进行特征空间降噪的可行性，试验结果表明所提方法可有效的降低计算时间，且明显提高机械运行状态识别及聚类精度，提高流形学习算法在机械故障诊断的适用性；
(2)针对局部线性嵌入算法LLE(Local Linear Embedding)中近邻点数选择对降维效果影响非常敏感的问题，通过分析得出不同样本的最优近邻点数应该不相等的结论，进而提出了可变近邻的LLE算法，提高了算法的聚类效果。将LLE算法的泛化形式： NPE(Neighborhood Preserving Embedding)算法与自组织映射SOM(Self-Organizing Map)结合，实现轴承退化过程的状态识别；
(3)针对局部保持投影LPP(Locality Preserving Projection)算法只考虑样本邻域信息而忽略距离较远样本信息的问题，提出同时考虑样本近邻信息及最远样本信息的保持投影算法：NFDPP(Nearest-Farthest Distantce Preserving Projection)，更好的保留数据结构的有效信息。发动机失火实验及轴承障实验结果表明，所提算法可有效提高机械运行状态的识别正确率；
(4)针对谱回归算法SR(Spectral Regression)未综合考虑样本局部及全局信息的问题，提出了同时考虑局部结构和全局数据结构的谱回归分析算法(Local and GlobalSpectral Regression, LGSR)。发动机实验及变速器故障实验表明，改进的谱回归算法能够获得更高的识别精度和聚类效果；
(5)针对多传感器测量系统，在前述研究基础上分别提出多维度的NFDPP(Multi-NFDPP)与多维度LGSR算法(Multi-LGSR)。将算法分别应用于多传感器监测系统的齿轮故障检测及轴承退化过程的在线监测，结果表明，这些方法能够有效的预测故障的发生并确定故障出现的部位。
Mechanical equipment is a complicated nonlinear system. Vibration signal analysis isone of the common methods for fault diagnosis, as it contains rich information to describethe mechanical running conditions. The signals are measured from the key parts of amachine, and then features are extracted to represent the running status. The widely usedvibration signal processing methods can be divided into: time domain analysis, frequencydomain analysis, time-frequency domain analysis. Due to the system complexity, the signalis always nonlinear and non-gaussian, which makes the fault diagnosis more difficult. It isnecessary to find new fault diagnosis method for complex nonlinear system diagnosis,keeping the accidents from the beginning and reducing the loss.
Manifold learning is one of the most popular focuses in pattern recognition, which isto represent the nonlinear data structure by a nonlinear map from the high-dimension spaceto a low-dimension space, and remaining the most useful information in the subspacesimultaneously. Therefore, we can adopt manifold learning to process various featuresextracted from time domain, frequency domain, time-frequency domain or to analysismulti-source sensor signals to recognize the machine running states. However, there arestill some problems when using manifold learning:(1) noise influences the robustness ofmanifold learning;(2) the parameter affects the mapping result;(3) some methods cannotpreserve the most useful information. So, based on the basics of manifold learning, weinvestigated different manifold learning algorithms in machine condition recognition andperformance assessment, with regard to the noise effect. The main research is as follows:
(1)It is time-consuming and memory-consuming to de-noise time signal traditionally,which also leads to difficulties in real time diagnosis. Based on the feature analysis, thenoise contained in the vibration data is transferred to the features. Thus, we de-noise thesefeatures directly to enhance the computational efficiency and conserve the memoryrequirements, which is beneficial to the application of manifold learning in machine faultdiagnosis;
(2)Since Locally Linear Embedding(LLE) is very sensitive to the numbers of nearestneighbor, which affects the dimension reduction. Based on the analysis of sample neighborselection, we propose a variable k-nearest neighbor locally linear embedding (VKLLE)algorithm to improve the classification and stability. NPE (Neighborhood PreservingEmbedding) is a linear approximation of the LLE, which is developed for out-of-sample problem. In this paper, NPE and SOM (Self-Organizing Map) are combined to assess thebearing degradation performance;
(3)The LPP only focus on local neighborhood information, which neglects the otherfurther samples. A novel method named NFDPP (Nearest-Farthest Distantce PreservingProjection) is proposed to explore data structure by considering a sample’s nearestneighbors and farthest samples at the same time. The experiment results of thebearing-defect classification and engine-fault diagnosis validate that the proposed NFDPPapproach achieves the good performance;
(4)Because the SR (Spectral Regression) does not take the global structure intoaccount, a novel feature extraction algorithm, called local and global spectralregression(LGSR), is presented for fault feature extraction. Gear and engine faultexperiments results demonstrate that the LGSR can extract identity information formachine defect classification;
(5)Based on the research on NFDPP and LGSR, two multi-way data processingalgorithms, denoted as the Multi-NFDPP and Multi-LGSR, are presented for processingmulti-sensor signals. The Gearbox fault detection experiments and bearing degradationassessment indicate that the proposed algorithms can effectively predict the occurrence ofdefect and find the defect location.
流形学习算法是近年来模式识别领域的研究热点之一，其本质在于通过一定的非线性映射将高维空间的数据结构在低维空间中进行表示，同时最大程度的保留高维空间数据的有用信息。因此，可以利用流形学习的优点，用以处理从时域、频域以及时频域中提取的多维信号特征或者处理由多个传感器获取的多源信号，实现机械运行状态的识别。然而，研究中发现存在以下问题：(1)噪声直接影响着流形学习算法的稳健性；(2)参数选择影响了算法的特征提取效果；(3)某些算法对高维数据结构信息保持不完整等。为此，本文从流形学习算法的基础理论出发，研究算法在机械系统状态识别、趋势分析中的噪声敏感性以及参数选择等问题，具体工作如下：
(1)传统时域降噪方法需要消耗大量的计算时间及存储空间，不利于实现机械系统在线诊断。提出直接对特征样本空间进行降噪的方法，理论分析了进行特征空间降噪的可行性，试验结果表明所提方法可有效的降低计算时间，且明显提高机械运行状态识别及聚类精度，提高流形学习算法在机械故障诊断的适用性；
(2)针对局部线性嵌入算法LLE(Local Linear Embedding)中近邻点数选择对降维效果影响非常敏感的问题，通过分析得出不同样本的最优近邻点数应该不相等的结论，进而提出了可变近邻的LLE算法，提高了算法的聚类效果。将LLE算法的泛化形式： NPE(Neighborhood Preserving Embedding)算法与自组织映射SOM(Self-Organizing Map)结合，实现轴承退化过程的状态识别；
(3)针对局部保持投影LPP(Locality Preserving Projection)算法只考虑样本邻域信息而忽略距离较远样本信息的问题，提出同时考虑样本近邻信息及最远样本信息的保持投影算法：NFDPP(Nearest-Farthest Distantce Preserving Projection)，更好的保留数据结构的有效信息。发动机失火实验及轴承障实验结果表明，所提算法可有效提高机械运行状态的识别正确率；
(4)针对谱回归算法SR(Spectral Regression)未综合考虑样本局部及全局信息的问题，提出了同时考虑局部结构和全局数据结构的谱回归分析算法(Local and GlobalSpectral Regression, LGSR)。发动机实验及变速器故障实验表明，改进的谱回归算法能够获得更高的识别精度和聚类效果；
(5)针对多传感器测量系统，在前述研究基础上分别提出多维度的NFDPP(Multi-NFDPP)与多维度LGSR算法(Multi-LGSR)。将算法分别应用于多传感器监测系统的齿轮故障检测及轴承退化过程的在线监测，结果表明，这些方法能够有效的预测故障的发生并确定故障出现的部位。
Mechanical equipment is a complicated nonlinear system. Vibration signal analysis isone of the common methods for fault diagnosis, as it contains rich information to describethe mechanical running conditions. The signals are measured from the key parts of amachine, and then features are extracted to represent the running status. The widely usedvibration signal processing methods can be divided into: time domain analysis, frequencydomain analysis, time-frequency domain analysis. Due to the system complexity, the signalis always nonlinear and non-gaussian, which makes the fault diagnosis more difficult. It isnecessary to find new fault diagnosis method for complex nonlinear system diagnosis,keeping the accidents from the beginning and reducing the loss.
Manifold learning is one of the most popular focuses in pattern recognition, which isto represent the nonlinear data structure by a nonlinear map from the high-dimension spaceto a low-dimension space, and remaining the most useful information in the subspacesimultaneously. Therefore, we can adopt manifold learning to process various featuresextracted from time domain, frequency domain, time-frequency domain or to analysismulti-source sensor signals to recognize the machine running states. However, there arestill some problems when using manifold learning:(1) noise influences the robustness ofmanifold learning;(2) the parameter affects the mapping result;(3) some methods cannotpreserve the most useful information. So, based on the basics of manifold learning, weinvestigated different manifold learning algorithms in machine condition recognition andperformance assessment, with regard to the noise effect. The main research is as follows:
(1)It is time-consuming and memory-consuming to de-noise time signal traditionally,which also leads to difficulties in real time diagnosis. Based on the feature analysis, thenoise contained in the vibration data is transferred to the features. Thus, we de-noise thesefeatures directly to enhance the computational efficiency and conserve the memoryrequirements, which is beneficial to the application of manifold learning in machine faultdiagnosis;
(2)Since Locally Linear Embedding(LLE) is very sensitive to the numbers of nearestneighbor, which affects the dimension reduction. Based on the analysis of sample neighborselection, we propose a variable k-nearest neighbor locally linear embedding (VKLLE)algorithm to improve the classification and stability. NPE (Neighborhood PreservingEmbedding) is a linear approximation of the LLE, which is developed for out-of-sample problem. In this paper, NPE and SOM (Self-Organizing Map) are combined to assess thebearing degradation performance;
(3)The LPP only focus on local neighborhood information, which neglects the otherfurther samples. A novel method named NFDPP (Nearest-Farthest Distantce PreservingProjection) is proposed to explore data structure by considering a sample’s nearestneighbors and farthest samples at the same time. The experiment results of thebearing-defect classification and engine-fault diagnosis validate that the proposed NFDPPapproach achieves the good performance;
(4)Because the SR (Spectral Regression) does not take the global structure intoaccount, a novel feature extraction algorithm, called local and global spectralregression(LGSR), is presented for fault feature extraction. Gear and engine faultexperiments results demonstrate that the LGSR can extract identity information formachine defect classification;
(5)Based on the research on NFDPP and LGSR, two multi-way data processingalgorithms, denoted as the Multi-NFDPP and Multi-LGSR, are presented for processingmulti-sensor signals. The Gearbox fault detection experiments and bearing degradationassessment indicate that the proposed algorithms can effectively predict the occurrence ofdefect and find the defect location.
引文
[1]李兵,陈雪峰,向家伟,等.基于小波有限元法的悬臂梁裂纹识别的试验研究[J].机械工程学报,2005,5(41):114-118.
[2]22年生涯走过28次太空之旅哥伦比亚号在返航途中失事[J].中国航天,2003,03:4-7
[3]潘旭峰,李晓雷,耿立恩.汽车传动系统机械故障诊断方法的研究[J].汽车工程,1998,3:166-164.
[4]赵文华,沈岩,陈黎明.电弧加热发动机羽流的非平衡态光谱诊断[J].工程热物理学报,2004,25(s):225-228.
[5]李岳,温熙森,吕克洪.基于核主成分分析的铁谱磨粒特征提取方法研究[J].国防科技大学学报,2007,29(2):113-118.
[6] PLATZ R, et al. Model Based Unbalance and Fatigue Crack Identification in Rotor Systems[C].EUROMECH Colloquium473on Identification and Updating Methods of MechanicalStructures, Prague Czech Republic,2002.
[7]胡守仁.神经网络导论[D].长沙:国防科技大学出版社,1993.
[8] Platt J. Sequential minimal optimization for SVM[OL]. http://www.ics.uci.edu/～xge/svm/Smo.html,1998.
[9]杜卫华,刘晓颖.基于人工智能方法的复杂过程故障诊断技术[J].控制工程,2002,9(4):1-6.
[10]焦李成.神经网络系统理论[M].西安:电子科技大学出版社,1990.
[11] Wei Qiao, Jiaqi Liang, Ganesh K,et al. Computational Intelligence for Control of Wind TurbineGenerators[J].2011IEEE Power and Energy Society General Meeting:1-6.
[12] Zijun Zhang, Anoop Verma, Andrew Kusiak. Fault Analysis and Condition Monitoring of theWind Turbine Gearbox[J]. IEEE TRANSACTIONS ON ENERGY CONVERSION,2012,2(27):526-535.
[13] Andrew Kusiak, Anoop Verma. A Data-Driven Approach for Monitoring Blade Pitch Faults inwind Turbines[J]. IEEE TRANSACTIONS ON SUSTAINABLE ENERGY,2011,1(2):87-96.
[14] Mitchell Lebold,Kenneth Maynard,et.al.Technology Development For Shaft CrackDetection in Rotating Equipment[C]. EPRI International Maintenance Conference, EPRIInternational Maintenance Conference, Chicago, Illinois,2003, August18-20.
[15] Kacprzynski G J, Roemer M J, et al. Enhancement of Physics-of-Failure Prognostic Modelswith System Level Features[J]. IEEE Aerospace Conference, Big Sky, MT,2002:9-16.
[16] Gang Cheng,Yu-long Cheng,Li-hua Shen, et al. Gear fault identification based onHilbert–Huang transform and SOM neural network[J]. Measurement,2013,46:1137–1146.
[17] Haifeng Tang, Jin Chen, Guangming Dong. Sparse representation based latent componentsanalysis for machinery weak fault detection[J]. Mechanical Systems and Signal Processing, InPress, Corrected Proof.
[18] Yaguo Lei, Dong Han, Jing Lin, et al. Planetary gearbox fault diagnosis using an adaptivestochastic resonance method[J]. Mechanical Systems and Signal Processing,2013,38:113–124.
[19] Yaguo Lei, Jing Lin, Zhengjia He, et al. A method based on multi-sensor data fusion for faultdetection of planetary gearboxes[J]. Sensors,2012(12):2005-2017.
[20]张周锁，闫晓旭，成玮.粒计算及其在机械故障智能诊断中的应用[J].西安交通大学学报，2009,43(9):37-41.
[21]李学军,杨大炼,郭灯塔,等.基于基座多传感核主元分析的故障诊断[J].仪器仪表学报,2011,(32):1551-1557.
[22]刘雨,陈进,潘玉娜,等.基于SVDD与信息融合技术的设备性能退化评估[J].振动与冲击,2009,(28):21-24.
[23]胡雷,胡茑庆,秦国军.双阈值单类支持矢量机在线故障检测算法及应用[J].机械工程学报，2009,45(3):169-173.
[24]李志农,范涛,刘立州,等.基于变分贝叶斯理论的机械故障源盲分离方法研究[J].振动与冲击,2009,28(6):12-16.
[25]于湘涛,褚福磊,郝如江.基于柔性形态滤波和支持矢量机的滚动轴承故障诊断方法[J].机械工程学报,2009,45(7):75-80.
[26]徐增丙,轩建平,史铁林,等.基于小波灰度矩向量与连续马尔可夫模型的轴承故障诊断[J].中国机械工程,2008,19(15):1858-1862.
[27]陶新民,徐晶,付强,等.基于样本密度KFCM新算法及其在故障诊断的应用[J].振动与冲击,2009,28(8):61-64.
[28]魏少华.基于声强知识与神经网络融合技术的发动机故障诊断研究[D].武汉:南京理工大学博士学位论文,2006.12.
[29]王志华.基于模式识别的柴油机故障诊断技术研究[D].武汉:武汉理工大学博士学位论文,2004.
[30]李宏坤.基于信息融合技术船舶柴油机故障诊断方法的研究与应用[D].大连:大连理工大学博士学位论文,2003.
[31]孙明明.流形学习理论与算法研究[D].南京:南京理工大学博士学位论文,2007.
[32] Bregler C, Omohundro S.M. Nonlinear manifold leaming for visual speech Recognition[C]. Int.Conf. Computer Vision,1995.
[33] Bregler C, Omohundro S.M. Nonlinear image interpolation using manifold leaming[C].Advanced in Neural Information Proeessing Systems&,MITPress,1995.
[34] H.S.Seung, D.D. Lee. The manifold ways of perception [J]. Science,2000,290:2268-2269.
[35] TENENBAUM J B, SILVA V D, LANGFORD J C.A global geometric framework for nonlineardimensionality reduction[J]. Science,2000,290:2319-2323.
[36] ROWEIS S T,SAUL L K.Nonlinear dimensionality reduction by locally linear embedding[J]. Science,2000,290:2323-2326.
[37] JOLLIFFE I.Principal component analysis[M].New York:Springer,1986.
[38] E. Oja, A. Hyvarinen, J. Karhunen. Independent Component Analysis[J]John Wiley&Sons,2001.
[39] P.N.Belhumeour, J.P.Hespanha, D.J.Kriegman.Eigenface vs. Fisherfaces: Recognition UsingClass Specific Linear Projection [J].Pattern Analysis and Machine Intelligence,1997,19(7):711-720.
[40] Belkin M,Niyogi P. Laplacian eigenmaps for dimensionality reduction and data representation[J].Neural Computation,2003,15(6),1373-1396.
[41] Zhang, Z.Y, Zha, H.Y. Principal manifolds and nonlinear dimensionality reduction via tangentspace alignment[J]. SIAM J. Sci. Comput,2004,26(1):313–338.
[42] X. He, D. Cai, S. Yan, H. Zhang, Neighborhood preserving embedding[J]. Proceedings inInternational Conference on Computer Vision (ICCV),2005.
[43] X. He and P. Niyogi, Locality Preserving Projections[J]. Advances in Neural InformationProcessing Systems16, Vancouver, British Columbia, Canada,2003.
[44] Zhenyue Zhang, Jing Wang, Hongyuan Zha. Adaptive Manifold Learning[J]. IEEE transactionson pattern analysis and machine intelligence,2012,34(22):253-265.
[45] Xiao B, Yu H, Hancock E. Graph matching using manifold embedding [C]. Lecture Notes inComputer Science,2004,3211:352-359.
[46] Law M.H.C, Zhang N, Jain A.K.Nonlinear manifold learning for data stream [J]. Proceedings of SIAM Data Mining2004, Orlando, Florida.
[47]谭璐,易东云.基于非线性特征提取的图像识别[J].计算机工程,2005,31(13):54-55.
[48]袁远,季星来,孙之荣,等. Isomap在基因表达谱数据聚类分析中的应用[J].清华大学学报(自然科学版),2004,44(9):1286-1289.
[49] D. Cai, X. He, J. Han.Isometric projection[C]. Proceedings of AAAI Conference on ArtificialIntelligence,2007.
[50] Cai, D. Spectral regression: A regression framework for efficient regularized subspace learning.
[D].University of Illinois at Urbana-Champaign, Urbana, IL, USA,2009.
[51] Qingbo He, Xiangxiang Wang. Time–frequency manifold correlation matching for periodic faultidentification in rotating machines [J]. Journal of Sound and Vibration,2013,(332):2611–2626.
[52] Qingbo He. Time–frequency manifold for nonlinear feature extraction in machinery faultdiagnosis [J]. Mechanical Systems and Signal Processing,2013,(35):200–218.
[53] Qingbo He, Yongbin Liu, Qian Long, et al. Time-Frequency Manifold as a Signature forMachine Health Diagnosis[J]. IEEE TRANSACTIONS ON INSTRUMENTATION ANDMEASUREMENT,2012(61):1218-1230.
[54]栗茂林,王孙安,梁霖.利用非线性流形学习的轴承早期故障特征提取方法[J].西安交通大学学报,2010,44(5):45-49.
[55]徐金梧,吕勇,王海峰.局部投影算法及其在非线性时间序列分析中的应用[J].机械工程学报,2003,39(9):146-150.
[56]阳建宏,徐金梧,杨德斌,等.基于主流形识别的非线性时间序列降噪方法及其在故障诊断中的应用[J].机械工程学报,2006,42(8):154-158.
[57]王广斌.基于流形学习的旋转机械故障诊断方法研究[D].湖南:中南大学博士学位论文,2010
[58]梁霖,徐光华,栗茂林,等.冲击故障特征提取的非线性流形学习方法[J].西安交通大学学报,2009,11(23):95-99.
[59]王善鹏.基于流形学习的滚动轴承故障特征提取方法研究[D].大连:大连理工大学硕士论文,2013.
[60]刘丽娟,陈果,郝腾飞.基于流形学习与一类支持向量机的滚动轴承早期故障识别方法[J].中国机械工程,2013,5(24):628-633.
[61]李城梁,王仲生,姜洪开,等.自适应Hessian LLE在机械故障特征提取中的应用[J].振动工程学报,2013,5(26):758-763.
[62]张赟,李本威.基于最大方差展开的非线性信号降噪方法及其在故障诊断中的应用[J].中国科学,2010,8(40):940-945
[63]王雷.基于流形学习的滚动轴承故障诊断若干方法研究[D].大连:大连理工大学博士学位论文,2013
[64] Yixiang Huang, Xuan F. Zha, Jay Lee, et al. Discriminant diffusion maps analysis: A robustmanifold learner for dimensionality reduction and its applications in machine conditionmonitoring and fault diagnosis[J]. Mechanical Systems and Signal Processing,2013,(34):277–297.
[65] Daniel Pérez, Francisco J, García-Fernández, et al. Visual analysis of a cold rolling processusing a dimensionality reduction approach[J]. Engineering Applications of Artifcial Intelligence,2013,(26):1865–1871.
[66] Benwei Li, Yun Zhang. Supervised locally linear embedding projection (SLLEP) for machineryfault diagnosis[J].Mechanical Systems and Signal Processing,2011,25:3125–3134
[67] G.F.Wang, Y.W. Yang, Y.C. Zhang,et al. Vibration sensor based tool condition monitoringusing v-support vector machine and locality preserving projection[J].Sensors and Actuators A,2014,209:24–32
[68] Gang Rong, Su-Yu Liu, Ji-Dong Shao. Dynamic fault diagnosis using extended matrix andtensor locality preserving discriminant analysis[J]. Chemometrics and Intelligent LaboratorySystems,2012(116):41–46.
[69] Chao Liu, Dongxiang Jiang, Wenguang Yang. Global geometric similarity scheme for featureselection in fault diagnosis [J]. Expert Systems with Applications,2014(41):3585–3595.
[70] Mingbo Zhao, Xiaohang Jina, Zhao Zhang, et al. Fault diagnosis of rolling element bearings viadiscriminative subspace learning: Visualization and classifcation [J].Expert Systems withApplications,2014(41):3391–3401.
[71] Li Jiang, Jianping Xuan, Tielin Shi. Feature extraction based on semi-supervised kernelMarginal Fisher analysis and its application in bearing fault diagnosis [J]. Mechanical Systemsand Signal Processing,2013,(41):113–126.
[72] Zuqiang Su, Baoping Tang, Jinghua Ma, et al. Fault diagnosis method based on incrementalenhanced supervised locally linear embedding and adaptive nearest neighbor classifier[J].Measurement,2014(48):136–148.
[73] Baoping Tang, Tao Song, Feng Li, et al. Fault diagnosis for a wind turbine transmission systembased on manifold learning and Shannon wavelet support vector machine [J]. Renewable Energy,2014,(62):1-9.
[74]成忠,诸爱士,陈德钊.ISOMAP-LDA方法用于化工过程故障诊断[J].化工学报,2009,60(1):122-126.
[75]张沐光,宋执环.LPMVP算法及其在故障检测中的应用[J].自动化学报,2009,6(6):7676-772.
[76]蒋全胜,贾民平,胡建中,等.基于拉普拉斯特征映射的故障模式识别方法[J].系统仿真学报,2008,20(20):5710-5713.
[77] Quansheng Jiang, Minping Jia, Jianzhong Hu, et al. Machinery fault diagnosis using supervisedmanifold learning[J]. Mechanical Systems and Signal Processing,2009,23:2301–2311.
[78]夏鲁瑞,胡茑庆,秦国军.基于流形学习的涡轮泵海量数据异常识别算法[J].航空动力学报,2011,3(26):698-703
[79]黎敏,阳建宏,徐金梧等.基于高维空间流形变化的设备状态趋势分析方法[J].机械工程学报,2009,45(02):213-218.
[80] Min Li,Jinwu Xua, Jianhong Yang, et al. Multiple manifolds analysis and its application to faultdiagnosis[J]. Mechanical Systems and Signal Processing,2009,23:2500-2509
[81] Benkedjouhb, K. Medjaher, N, Zerhouni, et al. Remaining useful life estimation based onnonlinear feature reduction and support vector regression[J]. Engineering Applications ofArtifcial Intelligence,2013,(26):1751–1760.
[82] Jianbo Yu. A nonlinear probabilistic method and contribution analysis for machine conditionmonitoring[J]. Mechanical Systems and Signal Processing,2013,(37):293–314.
[83] Jianbo Yu. Local and Nonlocal Preserving Projection for Bearing Defect Classification andPerformance Assessment[J]. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,2012.(59):2363-2376.
[84] Alexander Bleakie, Dragan Djurdjanovic. Feature extraction, condition monitoring, and faultmodeling in semiconductor manufacturing systems[J].Computers in Industry2013,64:203–213
[85] Ji-Dong Shao, Gang Rong.Nonlinear process monitoring based on maximum variance unfoldingprojections[J].Expert Systems with Applications,2009,36:11332-11340
[86]张熠卓，徐光华，梁霖等.利用增量式非线性流形学习的状态监测方法[J].西安交通大学学报,2011,1(45):64-68
[87]张妮,田学民.基于等距离映射的非线性动态故障检测方法[J].上海交通大学学报,2011,8(45):1202-1206
[88]王健,冯健,韩志艳.基于流形学习的局部保持PCA算法在故障检测中的应用[J].控制与决策,2013,5(28):683-687
[89]曾恒.流形学习在高速列车安全性态评估数据分析中的应用[D].成都:西南交通大学硕士学位论文,2013.
[90]丁康,李巍华,朱小勇.齿轮及齿轮箱故障诊断实用技术[M].北京:机械工业出版社,2005.
[91]林慧斌.离散频谱校正理论的抗噪性能研究及其在工程中的应用[D].广州:华南理工大学博士学位论文,2010.
[92]边肇祺,张学工.模式识别[M].2版.北京：清华大学出版社,2000.
[93] W. Xu,X. Liu,Y. Gong. Document Clustering Based on Non-negative Matrix Factorization[C].ACM SIGIR Conference on Information Retrieval,2003.
[94] Choon-Su Park,Young-Chul Choi, Yang-Hann Kim. Early fault detection in automotive ballbearings using the minimum variance cepstrum[J]. Mechanical Systems and Signal Processing,2013,38:534–548.
[95] Wenliao Du,Yanming Li,Jin Yuan,et al. Denoising with advanced stepwise false discovery ratecontrol and its application to fault diagnosis[J]. Measurement,2012,45:1515-1526.
[96]林京,屈梁生.基于连续小波变换的奇异性检测与故障诊断[J].振动工程学报,2000,13(4):523-529.
[97]吕志民,张军武,徐金梧.基于奇异谱的降噪方法及其在故障诊断技术中的运用[J].机械工程学报,1999,35(3):85-88.
[98] N.Sawali, R.B. Randall. Simulating gear and bearing interactions in the presence of faultsPart1. The combined gear bearing dynamic model and the simulation of localized bearingfaults[J]. Mechanical Systems and Signal Processing,2008,22:1924-1951.
[99] PAN Yaozhang, GE S S, MAMUM A A.Weighted locally linear embedding for dimensionreduction[J]. Pattern Recognition,2009,42(5):798-811.
[100]文贵华,江丽君,文军.邻域参数动态变化的局部线性嵌入[J].软件学报,2008,19(7):1666-1673.
[101]惠康华,肖柏华,王春恒.基于自适应近邻参数的局部线性嵌入[J].模式识别与人工智能，2010,23(6):842-846.
[102] KOUROPTEVA O，OKUN O，PIETIKLLINEN M.Selection of the optimal parameter valuefor the locally linear embedding algorithm [C].Proc of the1st International Conference onFuzzy Systems and Knowledge Discovery,Singapore,2002:359-363.
[103]王和勇,郑杰,姚正安,等.基于聚类和改进距离的LLE方法在数据降维中的应用[J].计算机研究与发展,2006,43(8):1485-1490.
[104]李欣鑫.基于核函数主元分析的半监督故障分类方法研究[D].广州:华南理工大学,2007.
[105] ROWEIS S.Data for MATLAB hackers [DB/OL].[2010-01-12].http://www.cs.nyu.edu/~roweis/data.html.
[106] Kohonen T, Self-Organizing Maps[J].Third extended edition, Springer,2001.
[107] Liao G, Liu S, Shi T, Zhang G. Gearbox Condition Monitoring Using Self-Organizing FeatureMaps [J]. Journal of Mechanical Engineering Science,2004,218:119-129.
[108] Hai Qiu, Jay Lee, Jing Lin, et.al. Robust performance degradation assessment methods forenhanced rolling element bearing prognostics[J]. Advanced Engineering Informatics,2003,17:127-140.
[109] Loparo.K.A et al., Bearingdata center, Case Western Reserve University.http://csegroups.case.edu/bearingdatacenter/pages/welcome-case-western-reserve-university-bearing-data-center-website, Retrieved,2013.
[110] E. Levina, P.J. Bickel. Maximum likelihood estimation of intrinsic dimension[J].In Advances inNeural Information Processing Systems,17, Cambridge, MA, USA, MIT Press,2004.
[111] Zhanguo Xia, Shixiong Xia, Ling Wan, et al. Spectral regression based fault feature extractionfor bearing accelerometer sensor signals[J]. Sensors,2012,12:13694－13719.
[112] Nomikos, P, MacGregor,J.F.Multivariate SPC charts for monitoring batch processes[J].Technometrics,1995,37:41-59.
[113] Qiu H,Lee J,Lin J,et al.Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics[J]. Journal of Sound and Vibration,2006,289(4/5):1066-1090.
[2]22年生涯走过28次太空之旅哥伦比亚号在返航途中失事[J].中国航天,2003,03:4-7
[3]潘旭峰,李晓雷,耿立恩.汽车传动系统机械故障诊断方法的研究[J].汽车工程,1998,3:166-164.
[4]赵文华,沈岩,陈黎明.电弧加热发动机羽流的非平衡态光谱诊断[J].工程热物理学报,2004,25(s):225-228.
[5]李岳,温熙森,吕克洪.基于核主成分分析的铁谱磨粒特征提取方法研究[J].国防科技大学学报,2007,29(2):113-118.
[6] PLATZ R, et al. Model Based Unbalance and Fatigue Crack Identification in Rotor Systems[C].EUROMECH Colloquium473on Identification and Updating Methods of MechanicalStructures, Prague Czech Republic,2002.
[7]胡守仁.神经网络导论[D].长沙:国防科技大学出版社,1993.
[8] Platt J. Sequential minimal optimization for SVM[OL]. http://www.ics.uci.edu/～xge/svm/Smo.html,1998.
[9]杜卫华,刘晓颖.基于人工智能方法的复杂过程故障诊断技术[J].控制工程,2002,9(4):1-6.
[10]焦李成.神经网络系统理论[M].西安:电子科技大学出版社,1990.
[11] Wei Qiao, Jiaqi Liang, Ganesh K,et al. Computational Intelligence for Control of Wind TurbineGenerators[J].2011IEEE Power and Energy Society General Meeting:1-6.
[12] Zijun Zhang, Anoop Verma, Andrew Kusiak. Fault Analysis and Condition Monitoring of theWind Turbine Gearbox[J]. IEEE TRANSACTIONS ON ENERGY CONVERSION,2012,2(27):526-535.
[13] Andrew Kusiak, Anoop Verma. A Data-Driven Approach for Monitoring Blade Pitch Faults inwind Turbines[J]. IEEE TRANSACTIONS ON SUSTAINABLE ENERGY,2011,1(2):87-96.
[14] Mitchell Lebold,Kenneth Maynard,et.al.Technology Development For Shaft CrackDetection in Rotating Equipment[C]. EPRI International Maintenance Conference, EPRIInternational Maintenance Conference, Chicago, Illinois,2003, August18-20.
[15] Kacprzynski G J, Roemer M J, et al. Enhancement of Physics-of-Failure Prognostic Modelswith System Level Features[J]. IEEE Aerospace Conference, Big Sky, MT,2002:9-16.
[16] Gang Cheng,Yu-long Cheng,Li-hua Shen, et al. Gear fault identification based onHilbert–Huang transform and SOM neural network[J]. Measurement,2013,46:1137–1146.
[17] Haifeng Tang, Jin Chen, Guangming Dong. Sparse representation based latent componentsanalysis for machinery weak fault detection[J]. Mechanical Systems and Signal Processing, InPress, Corrected Proof.
[18] Yaguo Lei, Dong Han, Jing Lin, et al. Planetary gearbox fault diagnosis using an adaptivestochastic resonance method[J]. Mechanical Systems and Signal Processing,2013,38:113–124.
[19] Yaguo Lei, Jing Lin, Zhengjia He, et al. A method based on multi-sensor data fusion for faultdetection of planetary gearboxes[J]. Sensors,2012(12):2005-2017.
[20]张周锁，闫晓旭，成玮.粒计算及其在机械故障智能诊断中的应用[J].西安交通大学学报，2009,43(9):37-41.
[21]李学军,杨大炼,郭灯塔,等.基于基座多传感核主元分析的故障诊断[J].仪器仪表学报,2011,(32):1551-1557.
[22]刘雨,陈进,潘玉娜,等.基于SVDD与信息融合技术的设备性能退化评估[J].振动与冲击,2009,(28):21-24.
[23]胡雷,胡茑庆,秦国军.双阈值单类支持矢量机在线故障检测算法及应用[J].机械工程学报，2009,45(3):169-173.
[24]李志农,范涛,刘立州,等.基于变分贝叶斯理论的机械故障源盲分离方法研究[J].振动与冲击,2009,28(6):12-16.
[25]于湘涛,褚福磊,郝如江.基于柔性形态滤波和支持矢量机的滚动轴承故障诊断方法[J].机械工程学报,2009,45(7):75-80.
[26]徐增丙,轩建平,史铁林,等.基于小波灰度矩向量与连续马尔可夫模型的轴承故障诊断[J].中国机械工程,2008,19(15):1858-1862.
[27]陶新民,徐晶,付强,等.基于样本密度KFCM新算法及其在故障诊断的应用[J].振动与冲击,2009,28(8):61-64.
[28]魏少华.基于声强知识与神经网络融合技术的发动机故障诊断研究[D].武汉:南京理工大学博士学位论文,2006.12.
[29]王志华.基于模式识别的柴油机故障诊断技术研究[D].武汉:武汉理工大学博士学位论文,2004.
[30]李宏坤.基于信息融合技术船舶柴油机故障诊断方法的研究与应用[D].大连:大连理工大学博士学位论文,2003.
[31]孙明明.流形学习理论与算法研究[D].南京:南京理工大学博士学位论文,2007.
[32] Bregler C, Omohundro S.M. Nonlinear manifold leaming for visual speech Recognition[C]. Int.Conf. Computer Vision,1995.
[33] Bregler C, Omohundro S.M. Nonlinear image interpolation using manifold leaming[C].Advanced in Neural Information Proeessing Systems&,MITPress,1995.
[34] H.S.Seung, D.D. Lee. The manifold ways of perception [J]. Science,2000,290:2268-2269.
[35] TENENBAUM J B, SILVA V D, LANGFORD J C.A global geometric framework for nonlineardimensionality reduction[J]. Science,2000,290:2319-2323.
[36] ROWEIS S T,SAUL L K.Nonlinear dimensionality reduction by locally linear embedding[J]. Science,2000,290:2323-2326.
[37] JOLLIFFE I.Principal component analysis[M].New York:Springer,1986.
[38] E. Oja, A. Hyvarinen, J. Karhunen. Independent Component Analysis[J]John Wiley&Sons,2001.
[39] P.N.Belhumeour, J.P.Hespanha, D.J.Kriegman.Eigenface vs. Fisherfaces: Recognition UsingClass Specific Linear Projection [J].Pattern Analysis and Machine Intelligence,1997,19(7):711-720.
[40] Belkin M,Niyogi P. Laplacian eigenmaps for dimensionality reduction and data representation[J].Neural Computation,2003,15(6),1373-1396.
[41] Zhang, Z.Y, Zha, H.Y. Principal manifolds and nonlinear dimensionality reduction via tangentspace alignment[J]. SIAM J. Sci. Comput,2004,26(1):313–338.
[42] X. He, D. Cai, S. Yan, H. Zhang, Neighborhood preserving embedding[J]. Proceedings inInternational Conference on Computer Vision (ICCV),2005.
[43] X. He and P. Niyogi, Locality Preserving Projections[J]. Advances in Neural InformationProcessing Systems16, Vancouver, British Columbia, Canada,2003.
[44] Zhenyue Zhang, Jing Wang, Hongyuan Zha. Adaptive Manifold Learning[J]. IEEE transactionson pattern analysis and machine intelligence,2012,34(22):253-265.
[45] Xiao B, Yu H, Hancock E. Graph matching using manifold embedding [C]. Lecture Notes inComputer Science,2004,3211:352-359.
[46] Law M.H.C, Zhang N, Jain A.K.Nonlinear manifold learning for data stream [J]. Proceedings of SIAM Data Mining2004, Orlando, Florida.
[47]谭璐,易东云.基于非线性特征提取的图像识别[J].计算机工程,2005,31(13):54-55.
[48]袁远,季星来,孙之荣,等. Isomap在基因表达谱数据聚类分析中的应用[J].清华大学学报(自然科学版),2004,44(9):1286-1289.
[49] D. Cai, X. He, J. Han.Isometric projection[C]. Proceedings of AAAI Conference on ArtificialIntelligence,2007.
[50] Cai, D. Spectral regression: A regression framework for efficient regularized subspace learning.
[D].University of Illinois at Urbana-Champaign, Urbana, IL, USA,2009.
[51] Qingbo He, Xiangxiang Wang. Time–frequency manifold correlation matching for periodic faultidentification in rotating machines [J]. Journal of Sound and Vibration,2013,(332):2611–2626.
[52] Qingbo He. Time–frequency manifold for nonlinear feature extraction in machinery faultdiagnosis [J]. Mechanical Systems and Signal Processing,2013,(35):200–218.
[53] Qingbo He, Yongbin Liu, Qian Long, et al. Time-Frequency Manifold as a Signature forMachine Health Diagnosis[J]. IEEE TRANSACTIONS ON INSTRUMENTATION ANDMEASUREMENT,2012(61):1218-1230.
[54]栗茂林,王孙安,梁霖.利用非线性流形学习的轴承早期故障特征提取方法[J].西安交通大学学报,2010,44(5):45-49.
[55]徐金梧,吕勇,王海峰.局部投影算法及其在非线性时间序列分析中的应用[J].机械工程学报,2003,39(9):146-150.
[56]阳建宏,徐金梧,杨德斌,等.基于主流形识别的非线性时间序列降噪方法及其在故障诊断中的应用[J].机械工程学报,2006,42(8):154-158.
[57]王广斌.基于流形学习的旋转机械故障诊断方法研究[D].湖南:中南大学博士学位论文,2010
[58]梁霖,徐光华,栗茂林,等.冲击故障特征提取的非线性流形学习方法[J].西安交通大学学报,2009,11(23):95-99.
[59]王善鹏.基于流形学习的滚动轴承故障特征提取方法研究[D].大连:大连理工大学硕士论文,2013.
[60]刘丽娟,陈果,郝腾飞.基于流形学习与一类支持向量机的滚动轴承早期故障识别方法[J].中国机械工程,2013,5(24):628-633.
[61]李城梁,王仲生,姜洪开,等.自适应Hessian LLE在机械故障特征提取中的应用[J].振动工程学报,2013,5(26):758-763.
[62]张赟,李本威.基于最大方差展开的非线性信号降噪方法及其在故障诊断中的应用[J].中国科学,2010,8(40):940-945
[63]王雷.基于流形学习的滚动轴承故障诊断若干方法研究[D].大连:大连理工大学博士学位论文,2013
[64] Yixiang Huang, Xuan F. Zha, Jay Lee, et al. Discriminant diffusion maps analysis: A robustmanifold learner for dimensionality reduction and its applications in machine conditionmonitoring and fault diagnosis[J]. Mechanical Systems and Signal Processing,2013,(34):277–297.
[65] Daniel Pérez, Francisco J, García-Fernández, et al. Visual analysis of a cold rolling processusing a dimensionality reduction approach[J]. Engineering Applications of Artifcial Intelligence,2013,(26):1865–1871.
[66] Benwei Li, Yun Zhang. Supervised locally linear embedding projection (SLLEP) for machineryfault diagnosis[J].Mechanical Systems and Signal Processing,2011,25:3125–3134
[67] G.F.Wang, Y.W. Yang, Y.C. Zhang,et al. Vibration sensor based tool condition monitoringusing v-support vector machine and locality preserving projection[J].Sensors and Actuators A,2014,209:24–32
[68] Gang Rong, Su-Yu Liu, Ji-Dong Shao. Dynamic fault diagnosis using extended matrix andtensor locality preserving discriminant analysis[J]. Chemometrics and Intelligent LaboratorySystems,2012(116):41–46.
[69] Chao Liu, Dongxiang Jiang, Wenguang Yang. Global geometric similarity scheme for featureselection in fault diagnosis [J]. Expert Systems with Applications,2014(41):3585–3595.
[70] Mingbo Zhao, Xiaohang Jina, Zhao Zhang, et al. Fault diagnosis of rolling element bearings viadiscriminative subspace learning: Visualization and classifcation [J].Expert Systems withApplications,2014(41):3391–3401.
[71] Li Jiang, Jianping Xuan, Tielin Shi. Feature extraction based on semi-supervised kernelMarginal Fisher analysis and its application in bearing fault diagnosis [J]. Mechanical Systemsand Signal Processing,2013,(41):113–126.
[72] Zuqiang Su, Baoping Tang, Jinghua Ma, et al. Fault diagnosis method based on incrementalenhanced supervised locally linear embedding and adaptive nearest neighbor classifier[J].Measurement,2014(48):136–148.
[73] Baoping Tang, Tao Song, Feng Li, et al. Fault diagnosis for a wind turbine transmission systembased on manifold learning and Shannon wavelet support vector machine [J]. Renewable Energy,2014,(62):1-9.
[74]成忠,诸爱士,陈德钊.ISOMAP-LDA方法用于化工过程故障诊断[J].化工学报,2009,60(1):122-126.
[75]张沐光,宋执环.LPMVP算法及其在故障检测中的应用[J].自动化学报,2009,6(6):7676-772.
[76]蒋全胜,贾民平,胡建中,等.基于拉普拉斯特征映射的故障模式识别方法[J].系统仿真学报,2008,20(20):5710-5713.
[77] Quansheng Jiang, Minping Jia, Jianzhong Hu, et al. Machinery fault diagnosis using supervisedmanifold learning[J]. Mechanical Systems and Signal Processing,2009,23:2301–2311.
[78]夏鲁瑞,胡茑庆,秦国军.基于流形学习的涡轮泵海量数据异常识别算法[J].航空动力学报,2011,3(26):698-703
[79]黎敏,阳建宏,徐金梧等.基于高维空间流形变化的设备状态趋势分析方法[J].机械工程学报,2009,45(02):213-218.
[80] Min Li,Jinwu Xua, Jianhong Yang, et al. Multiple manifolds analysis and its application to faultdiagnosis[J]. Mechanical Systems and Signal Processing,2009,23:2500-2509
[81] Benkedjouhb, K. Medjaher, N, Zerhouni, et al. Remaining useful life estimation based onnonlinear feature reduction and support vector regression[J]. Engineering Applications ofArtifcial Intelligence,2013,(26):1751–1760.
[82] Jianbo Yu. A nonlinear probabilistic method and contribution analysis for machine conditionmonitoring[J]. Mechanical Systems and Signal Processing,2013,(37):293–314.
[83] Jianbo Yu. Local and Nonlocal Preserving Projection for Bearing Defect Classification andPerformance Assessment[J]. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,2012.(59):2363-2376.
[84] Alexander Bleakie, Dragan Djurdjanovic. Feature extraction, condition monitoring, and faultmodeling in semiconductor manufacturing systems[J].Computers in Industry2013,64:203–213
[85] Ji-Dong Shao, Gang Rong.Nonlinear process monitoring based on maximum variance unfoldingprojections[J].Expert Systems with Applications,2009,36:11332-11340
[86]张熠卓，徐光华，梁霖等.利用增量式非线性流形学习的状态监测方法[J].西安交通大学学报,2011,1(45):64-68
[87]张妮,田学民.基于等距离映射的非线性动态故障检测方法[J].上海交通大学学报,2011,8(45):1202-1206
[88]王健,冯健,韩志艳.基于流形学习的局部保持PCA算法在故障检测中的应用[J].控制与决策,2013,5(28):683-687
[89]曾恒.流形学习在高速列车安全性态评估数据分析中的应用[D].成都:西南交通大学硕士学位论文,2013.
[90]丁康,李巍华,朱小勇.齿轮及齿轮箱故障诊断实用技术[M].北京:机械工业出版社,2005.
[91]林慧斌.离散频谱校正理论的抗噪性能研究及其在工程中的应用[D].广州:华南理工大学博士学位论文,2010.
[92]边肇祺,张学工.模式识别[M].2版.北京：清华大学出版社,2000.
[93] W. Xu,X. Liu,Y. Gong. Document Clustering Based on Non-negative Matrix Factorization[C].ACM SIGIR Conference on Information Retrieval,2003.
[94] Choon-Su Park,Young-Chul Choi, Yang-Hann Kim. Early fault detection in automotive ballbearings using the minimum variance cepstrum[J]. Mechanical Systems and Signal Processing,2013,38:534–548.
[95] Wenliao Du,Yanming Li,Jin Yuan,et al. Denoising with advanced stepwise false discovery ratecontrol and its application to fault diagnosis[J]. Measurement,2012,45:1515-1526.
[96]林京,屈梁生.基于连续小波变换的奇异性检测与故障诊断[J].振动工程学报,2000,13(4):523-529.
[97]吕志民,张军武,徐金梧.基于奇异谱的降噪方法及其在故障诊断技术中的运用[J].机械工程学报,1999,35(3):85-88.
[98] N.Sawali, R.B. Randall. Simulating gear and bearing interactions in the presence of faultsPart1. The combined gear bearing dynamic model and the simulation of localized bearingfaults[J]. Mechanical Systems and Signal Processing,2008,22:1924-1951.
[99] PAN Yaozhang, GE S S, MAMUM A A.Weighted locally linear embedding for dimensionreduction[J]. Pattern Recognition,2009,42(5):798-811.
[100]文贵华,江丽君,文军.邻域参数动态变化的局部线性嵌入[J].软件学报,2008,19(7):1666-1673.
[101]惠康华,肖柏华,王春恒.基于自适应近邻参数的局部线性嵌入[J].模式识别与人工智能，2010,23(6):842-846.
[102] KOUROPTEVA O，OKUN O，PIETIKLLINEN M.Selection of the optimal parameter valuefor the locally linear embedding algorithm [C].Proc of the1st International Conference onFuzzy Systems and Knowledge Discovery,Singapore,2002:359-363.
[103]王和勇,郑杰,姚正安,等.基于聚类和改进距离的LLE方法在数据降维中的应用[J].计算机研究与发展,2006,43(8):1485-1490.
[104]李欣鑫.基于核函数主元分析的半监督故障分类方法研究[D].广州:华南理工大学,2007.
[105] ROWEIS S.Data for MATLAB hackers [DB/OL].[2010-01-12].http://www.cs.nyu.edu/~roweis/data.html.
[106] Kohonen T, Self-Organizing Maps[J].Third extended edition, Springer,2001.
[107] Liao G, Liu S, Shi T, Zhang G. Gearbox Condition Monitoring Using Self-Organizing FeatureMaps [J]. Journal of Mechanical Engineering Science,2004,218:119-129.
[108] Hai Qiu, Jay Lee, Jing Lin, et.al. Robust performance degradation assessment methods forenhanced rolling element bearing prognostics[J]. Advanced Engineering Informatics,2003,17:127-140.
[109] Loparo.K.A et al., Bearingdata center, Case Western Reserve University.http://csegroups.case.edu/bearingdatacenter/pages/welcome-case-western-reserve-university-bearing-data-center-website, Retrieved,2013.
[110] E. Levina, P.J. Bickel. Maximum likelihood estimation of intrinsic dimension[J].In Advances inNeural Information Processing Systems,17, Cambridge, MA, USA, MIT Press,2004.
[111] Zhanguo Xia, Shixiong Xia, Ling Wan, et al. Spectral regression based fault feature extractionfor bearing accelerometer sensor signals[J]. Sensors,2012,12:13694－13719.
[112] Nomikos, P, MacGregor,J.F.Multivariate SPC charts for monitoring batch processes[J].Technometrics,1995,37:41-59.
[113] Qiu H,Lee J,Lin J,et al.Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics[J]. Journal of Sound and Vibration,2006,289(4/5):1066-1090.