本文目的主要是研究當旋轉機械發生零件鬆動時，使用HHT中的邊際譜方法來分析振動信號的可行性。通常機械的故障訊號都為非線性、非穩態的，所以利用傳統的傅立葉轉換對訊號做頻域分析，其頻譜無法真實地反映出非線性或非穩態振動的特質，所以必須採取其它方法來分析振動信號。因此我們採用黃鍔博士在1998年提出的希爾伯特黃轉換(Hilbert-Huang transform，HHT)，透過整體經驗模態分解法(Ensemble Empirical Mode Decomposition，EEMD)分解而得的內建模態函數(Intrinsic Mode Function，IMF)，並透過顯著性檢定(significance test)與除去低頻訊號，選取具有意義的內建模態函數(IMFs)來組成邊際譜，透過故障指標來分辨不同的振動信號。這個故障指標方法結合傳統馬氏距離(Mahalanobis distance)與餘弦指標(cosine)，將由使用邊際譜所得到的特徵頻率，根據這個新的距離辨別方法分析振動信號，將不同類型的鬆動故障特徵分辨出來，有效地的提高了機械鬆動的辨別率，並且經由圖形來說明HHT的優點。透過實驗室的轉子平台系統(PBS-5000型)模擬轉子零件鬆動的情形，根據分析振動訊號的結果顯示，本文提出新的故障診斷方法能夠提昇距離分量在不同的邊際譜分布，所以可以辨別不同類型的機械鬆動故障。

The purpose of the research in this paper is to investigate the feasibility of utilizing the marginal Hilbert spectra of vibration signals to identify the looseness faults at different components of rotating machinery. The faulty signals of rotating machinery are generally nonlinear and non-stationary. The traditional Fourier-based spectral analysis can not precisely describe the nonlinear or non-stationary vibration behavior. Therefore, it motivates this research to use an alternative method to analyze the complicated vibration signals. Under such conditions, Huang et al. (1998) propose a new method, termed as Hilbert-Huang Transform (HHT), to analyze the nonlinear or non-stationary data. Through the Ensemble Empirical Mode Decomposition (EEMD) method, the original data can be decomposed into several Intrinsic Mode Function (IMF) components. The information-contained IMFs are extracted through the significance test and removing the low frequency components. The marginal Hilbert spectra consist of the information-contained IMFs and represent the characteristics of the machine operating condition. The fault indicator index that is combined by Mahalanobis distance and cosine index is defined to measure the spectral similarities among different looseness types in the operating machinery. The effectiveness of the proposed method is evaluated by analyzing the marginal Hilbert spectra of vibration signals in the different types of looseness faults. A test bed of rotor-bearing system (PBS-5000) is performed to illustrate the looseness faults at different mechanical components. The results show that the proposed diagnose method is capable of enhancing the distinction quantities of different marginal Hilbert spectra distributions, and thus identify the type of looseness fault at machinery.

[1]Goldman, P. and Muszynska, A., “Application of Full Spectrum to Rotating Machinery Diagnostics,” Orbit, Vol. 20, No. 1, pp. 17-21, 1999.
[2]Nandi, S. and Toliyat, H. A., “Condition Motoring and Fault Diagnosis of Electrical Machines-A Review,” IEEE Industry Applications Conference, Vol. 1, pp. 197-204, 1999.
[3]Gayme, D., Menon, S., Ball, C., Mukavetz, D. and Nwadiogbu, E., “Fault detection and diagnosis in turbine engines using fuzzy logic,” Processing of International Conference on Fuzzy Information, North American, pp. 341-346, 2003.
[4]Mitchell, S. J., “Machinery Analysis and Monitoring,” PennWell Publishing Company, Oklahoma, 1981.
[5]Eshleman, R. L., “Machinery Vibration Analysis,” The Vibration Institute, VIPress, 1995.
[6]Eshleman, R. L. and Lewis, F. M., “Torsional Vibration in Reciprocating and Rotating Machines,” in Shock and Vibration Handbook, edutor Harris, C.M., McGraw-Hill, 1988.
[7]張嘉仁，“振動與噪音信號在引擎診斷上的應用”，國立台北科技大學/車輛工程系碩士論文，2002。
[8]Zhao, L. and Sheng, Z., “Combination of discrete cosine transform with neural network in fault diagnosis for rotating machine,” Proceedings of the IEEE international conference on industrial technology, pp. 450-454, 1996.
[9]趙安民，“馬達故障診斷之模糊類神經網路”，私立中原大學機械工程學系研究所碩士論文，2004。
[10]莊鵬騰，“階次追蹤分析技術應用於旋轉機器上之探討”，中正理工學院造船工程研究所碩士論文，1999。
[11]Wu, T. Y., Chung, Y. L. and Hong, H. C., “Looseness Fault Identification of Rotating Machinery through Analyzing Power Spectral Density Distribution of Vibration Signals,” The 32nd National Conference on Theoretical and Applied Mechanics, 2008.
[12]羅明哲，“小波轉換於轉子故障與切削顫振偵測之應用”，私立中原大學機械工程研究所碩士論文，2001。
[13]朱效賢，“包絡譜分析於軸承故障診斷之探討暨工程應用”，國立中央大學機械工程研究所碩士論文，2005。
[14]Barkov, A. and Barkova, N., “Condition assessment and life prediction of rolling element bearings-part 1,” S V Sound and Vibration, Vol. 29, No. 6, pp. 10-17, 1995.
[15]Finley, W. R. ,Hodowanec, M. M. and Holter, W. G., “An Analytical Approach to Solving Motor Vibration Problems,” IEEE Transactions on Industry Applications, Vol. 36, No. 5, pp. 1467-1480, 2000.
[16]Huang, N. E., Shen, Z., Long, S. R., Wu, M. C., Shih, H. H., Zheng, Q., Yen, N. C., Tung, C. C. and Liu H. H., “The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis,” Physical and Engineering Sciences London , A , No. 454, pp. 903-995, 1998.
[17]Peng, Z. K., Tse, P. W., Chu, F. L., “A comparison study of improved Hilbert-Hung Transform and Wavelet Transform: Application to Fault diagnosis for Rolling Bearing,” Mechanical Systems and signal Processing, Vol. 19, pp. 974-988, 2005.
[18]Yu, D., Cheng, J. and Yang, Y., “Application of EMD Method and Hilbert Spectrum to the Fault Diagnosis of Roller Bearings,” Mechanical Systems and signal Processing, Vol. 19, pp. 259-270, 2005.
[19]Oliver, J. A. and Kosmatka, J. B., “Evaluation of the Hilbert-Huang transform for modal-based structural health monitoring applications,” Proceedings of SPIE, Vol. 5767, pp. 274-285, 2005.
[20]于德介、程軍聖、楊宇，機械故障診斷的Hilbert-Huang 變換方法，科學出版社，2006。
[21]Wu, Z. and Huang, N. E., “Ensemble Empirical Mode Decomposition: a noise-assisted data analysis method,” Centre for Ocean-Land-Atmosphere Studies, Technical Report series, Vol. 193, No. 173, 2004.
[22]Cohen, L., “Time-frequency analysis,” Prentice-Hall, New York, 1995.
[23]Gabor, D., “Theory of communication,” Proceedings of the IEEE, Vol. 93, pp. 429-457, 1946.
[24]Tichmarsh, E. C., “Introduction to the theory of Fourier Integrals,” Oxford University Press, 1948.
[25]Newland, D. E., “An introduction to Random Vibrations, Spectral and Wavelet Analysis,” John Wiley and Sons, 1993.
[26]Long, S. R., Huang, N. E., Lung, C. C., Wu, M. L., Lin, R. Q., Mollo-Christensen, E. and Yuan, Y., “The Hilbert Techniques : An alternate approach for non-steady time series analysis,” IEEE Geoscience and Remote Sensing Society Letters, Vol. 3, pp. 6-11, 1995.
[27]Dazin, P. G., “Nonlinear Systems,” Cambridge University Press, Cambridge, 1992.
[28]Huang, N. E., Shen, Z. and Long, S. R., “A New View of Nonlinear Water Waves : The Hilbert Spectrum,” Annual Review of Fluid Mechanics, Vol. 31, pp. 417-457, 1999.
[29]Wu, Z. and Huang, N. E., “A study of the characteristics of white noise using the empirical mode decomposition method,” Proceedings of Royal Society London, A , No. 460, pp. 1597-1611, 2004.
[30]Wu, Z. and Huang, N. E., “Ensemble Empirical Mode Decomposition: a noise-assisted data analysis method,” Advances in Adaptive Data Analysis. Vol. 1, No. 1, pp .1-41, 2008.
[31]Mahalanobis, P. C., “On the generalised distance in statistics,” Proceedings of the National Institute of Sciences of India , Vol.12, pp. 49-55, 1936.
[32]Taguchi, G. and Jugulum, R., “The Mahalanobis-Taguchi Strategy–a Pattern Technology System,” JohnWiley and Sons, New York, 2002.