本研究提出一套有效系統應用於工具機性能之診斷，期望能即時有效的掌握工具機性能。為避免工具機過載運轉或損壞，工具機診斷系統可適時發出振動過大、過載溫升、異常噪音等預警訊息，並可診斷出各種異常狀態。首先經由三軸加速規、熱電耦、麥克風等感測器，擷取工具機各項物理信號，如振動、溫度、噪音等；再將信號資料，傳輸至本研究之工具機性能診斷系統，達到性能評估與診斷之功效。
本研究使用MEMS微型感測器擷取出工具機各項硬體設備之物理信號，首先利用模糊邏輯理論進行性能評估，判斷出各個信號之歸屬度，再透過歸屬度之分類定義出所屬之層級，判別出性能正常與否。倘若性能異常，則系統將判斷出異常項目為何。透過擷取振動與噪音物理信號，系統首先使用傅立葉轉換(Fourier Transform)分析各個信號在頻率域中的變化，再透過主成份分析(Principal Component Analysis)來降低資料維度；從信號資料的特徵當中，評估工具機的工作情況。將記錄後的信號資料，利用倒傳遞神經網路(BPN，Back Propagation Network)來建立特徵數學模型，再藉由誤差函數來判斷異常項目。最後將物理信號資料經由人機介面之診斷報告來完成工具機性能之診斷功能。透過工具機診斷系統的預警機制，提供維修運轉人員儘早因應，避免工具機因損壞而降低產值。

This study presents an effective system of performance diagnosis used in the diagnosis of machine tools, which expect to be immediately effective control machine tool performance. In this study, the machine tool performance diagnosis system can be timely to issue warning messages including excessive vibration, temperature overload, and extraordinary noise to avoid the machine tool operation overload or damage. First of all, by three-axis accelerometer, thermocouples, microphones and other sensors, the physical signal which is obtained by capture machine tools such as vibration, temperature, noise. Then transfer the signal data to the study of machine tool performance diagnosis system to achieve the effect of performance evaluation and diagnosis.
This study use MEMS sensors to extract the machine tools of the physical hardware signals. First, system uses the fuzzy logic theory in performance evaluation and determines the degree of ownership of all signals. Then through the classification of the membership defines the level of the categories to determine the performance which is normal or not. The system will determine abnormal items if the performance was abnormal.
First, analysis of each signal variation in the frequency domain by Fourier Transform, then reduce dimensionality through Principal Component Analysis. Assess the work of machine tools through the characteristics of the signal information. To establish the characteristic mathematical model from the signal data we recorded by Back Propagation Network. Then, determine the abnormal items through the percentage of error function. Finally, completed the function of diagnostics and forecast of performance of machine tool used the physical signal information through the diagnostic report of HMI. Provide the maintenance personnel as early as possible in response to operation through the early warning of diagnostic system of Machine tool to avoid the damage which to reduce the output of machine tool. Finally, completed the function of diagnostics and forecast of performance of machine tool used the physical signal information through the diagnostic report of HMI. Provide the maintenance personnel as early as possible in response to operation through the early warning of diagnostic system of Machine tool to avoid the damage which to reduce the output of machine tool.

[1] E.T. Whittaker, "On the Functions Which are Represented by the Expansions of the Interpolation Theory," Proc. Royal Soc. Edinburgh, Sec. A, vol.35, pp.181-194, 1915.
[2] J.W. Cooley and J.W. Tukey, "An Algorithm for the Machine Computation of the Complex Fourier Series," Mathematics of Computation, Vol. 19, pp. 297-301 , April 1965.
[3] B.F. Arnold, ”Staticstical Tests Optimally Certain Fuzzy Requirements on the Power Function and on the Sample Size”, Fuzzy Sets and System, Vol. 75, Issue 3,pp. 365-372, 1996.
[4] I.T. Jolliffe, “Principal Component Analysis”, Springer-Verlag. pp.487, 1986.
[5] Jian-Da Wu, Jun-Ming Kuo, “An automotive generator fault diagnosis system using discrete wavelet transform and artificial neural network” ,Expert Systems with Applications, Volume 36, Issue 6, Pages 9776-9783, August 2009.
[6] W. Li, D. Li, J. Ni ,”Diagnosis of tapping process using spindle motor current”, International Journal of Machine Tools and Manufacture, Volume 43, Issue 1, Pages 73-79, January 2003.
[7] P. Duhamel and M. Vetterli, "Fast Fourier Transforms: A Tutorial Review and a State of the Art," Signal Processing, Vol. 19, pp. 259-299, April 1990.
[8] N. Casoetto, D. Djurdjanovic ,“Multisensor Process Performance Assessment through Use of Autoregressive Modeling and Feature Maps”, Journal of Manufacturing Systems, Vol.22, No.1, pp.1-9, 2003.
[9] D. Djurdjanovic, J. Lee and Jun Ni, “Watchdog Agent An Infotronics Based Prognostics Approach for Product Performance Degradation Assessment and Prediction”, Elsevier, Vol.17, pp.109-125, 2003.
[10] Jae-Seob Kwak, Ji-Bok Song ,“Trouble diagnosis of the grinding process by using acoustic emission signals” ,International Journal of Machine Tools and Manufacture, Volume 41, Issue 6, Pages 899-913, May 2001.
[11] S. Dey, J. A. Stori,” A Bayesian network approach to root cause diagnosis of process variations”, International Journal of Machine Tools and Manufacture, Volume 45, Issue 1, Pages 75-91, January 2005.
[12] J. Sun, G.S. Hong, Y.S. Wong, M. Rahman, Z.G. Wang,” Effective training data selection in tool condition monitoring system”, International Journal of Machine Tools and Manufacture, Volume 46, Issue 2, Pages 218-224, February 2006.
[13] Martin B. Jun, O. Burak Ozdoganlar, Richard E. DeVor, Shiv G. Kapoor, Andreas Kirchheim, Georges Schaffner,” Evaluation of a spindle-based force sensor for monitoring and fault diagnosis of machining operations”, International Journal of Machine Tools and Manufacture, Volume 42, Issue 6, Pages 741-751, May 2002.
[14] Seung-Han Yang, Ki-Hoon Kim, Yong Kuk Park,“Measurement of spindle thermal errors in machine tool using hemispherical ball bar test”, International Journal of Machine Tools and Manufacture, Volume 44, Issues 2-3, Pages 333-340, February 2004.
[15] Paul R. Drake, Difu Pant ,“Multiple fault diagnosis for a machine tool's flood coolant system using a neural network”, International Journal of Machine Tools and Manufacture, Volume 36, Issue 11, Pages 1247-1251, November 1996.
[16] D. E. Dimla Jr, P. M. Lister, N. J. Leighton ,“Neural network solutions to the tool condition monitoring problem in metal cutting—A critical review of methods”, International Journal of Machine Tools and Manufacture, Volume 37, Issue 9, Pages 1219-1241, September 1997.
[17] 羅世杰，運轉機械振動狀態監控在工具機的應用，國立中興大學機械工程學系所碩士論文，台中，2008。
[18] 鍾宇濤，時頻分析於振動信號之故障診斷，中原大學機械工程研究系所碩士論文，中壢，2008。
[19] 劉憲正，具線上性能預測之智慧型機械系統，國立成功大學製造工程研究所碩士論文，2007。
[20] 吳宏亮，遠端監控系統應用於工具機切削動態異常診控之應用研究，中原大學機械工程所碩士論文，中壢，2006。
[21] 洪瑞廷，以小波包轉換與貝式分類法進行機械性能評估，國立成功大學製造工程研究所碩士論文，2007。
[22] 吳國安，應用RBF類神經網路於綜合加工機主軸溫升變形之預測。大葉大學機械工程所碩士論文，彰化，2005。
[23] 王智中，多解析盲蔽反捲積演算法應用於振動與聲音之研究，國立成功大學系統與船舶機電工程研究所博士論文，台南，2003。
[24] 趙安民，馬達故障診斷之模糊類神經網路，中原大學機械工程所碩士論文，中壢，2004。
[25] 黃嘉閔，轉動機械之智慧型階次追蹤診斷系統，交通大學機械工程研究所碩士論文，新竹，2003。
[26] 葉懷仁，類神經網路應用於引擎聲訊之辨識，國立中山大學物理研究所碩士論文，高雄，2003。
[27] 陳政哲，利用類神經網路於雷射切割QFN封裝之切割品質預測之研究，國立台灣科技大學自動化及控制研究所碩士論文，台北，2006。