本文所要探討的是利用高感度麥克風擷取引擎噪聲，進而經過電腦程式的處理，得到引擎各別的均方根值(Root Mean Square Value)、峰態值（Kurtosis）、頻譜分析圖、頻譜區間功率值以及頻譜區間比值等，來辨識情況未明引擎為正常抑或是故障引擎，接著由門檻值判斷後所產生之數字序列來判斷故障的種類。量測實驗在車廠廠房裡的引擎試台上所進行，當測試機台上之引擎暖機完畢後，直接將引擎轉速拉升到本實驗所訂定之轉速3000 rpm(轉/分)，等待引擎轉速穩定後，即進行量測。利用多具車廠確定為正常的引擎來收集聲音訊號，建立資料庫，並從資料庫中訂定選取判斷條件的門檻值，其後則由已確認故障狀況之引擎所擷取下來的訊號資料進行比對，並觀察是否有故障狀況特徵可以擷取。經實驗驗證，本論文所用實驗的方法能夠有效的判斷引擎的狀況。

In this paper, a high-sensitivity microphone was used in acquiring the sound pressure variations of engines. Then, some parameters such as, Root Mean Square values、Kurtosis values、spectrum analysis charts、spectrum interval power values and spectrum interval specific ratios are evaluated from computer program to distinguish whether unidentified engines to be regular or failure. Furthermore, the threshold values of these parameters for the classifications of failure types were determined by generating a series of digits.
Measurements were performed at the test site in an automobile engine plant. After finishing the warm-up procedure of the engine at site, the rotation rate was raised to 3000 rpm directly, and while the rotation rate reaches a steady state condition, the measurements were started. Then by collecting sound signals from several engines which were recognized to be regular by the automobile engine plant, a database is established, and the threshold values of the chosen parameters are determined. Finally, from the analysis of the collected signals, the characteristics of the failure engine could be obtained by comparing the differences between the regular and the failure engines.
The experimental data demonstrates that our method of experiment could diagnose the status of engines effectively.

[1]劉成群，張超群，〝汽車振動與噪音〞，新文京開發出版有限公司，p349~p367，2002。

[2]James W.Cooley and John W.Tukey，〝An algorithm for the machine calculation of complex Fourier series〞， Mathematical Computation，Vol 19，NO.90，p297~p301，1965。

[3]James S. Walker，〝Fast Fourier transforms〞，CRC Press，1996。

[4]D.Gabor，〝Theory of communication〞，J.IEE.93，p429~p457，1946。

[5]Dai F. and Sugisaka M.〝Fuzzy fault query approach in the fault diagnosis of a power plant system〞，Proceedings IEEE/RSJ International Conference on Intelligent Robots and Systems，Vol.2，1999，p752~p756。

[6]王智中，〝小波理論應用於旋轉機械故障診斷之分析〞，國立成功大學造船暨船舶機電工程學系碩士論文，1999。

[7]Wang X G. and Liu W.，〝A fuzzy fault diagnosis scheme with application〞，IFSA World Congress & 20th NAFIPS International Conference，Vol.3，2001，p1489~p1491。

[8]陳聖洲，〝汽油引擎之類神經模糊故障診斷〞，南台科技大學機械工程學系碩士論文，2002。

[9]高東正，〝統計學概要〞，科技圖書股份有限公司，1982，p217~p226。

[10]童甲春，〝統計學STATISTICS〞，前程企業管理有限公司，1997，p82~p85。

[11]Alan V.Oppenheim，Ronald W.Schafer，〝DISCRETE-TIME SIGNAL PROCESSING〞，Prentice Hall，Inc，1989，p144~p145。

[12]Thomas P.Krauss，Loren Shure，John N.Little〝Signal Processing Toolbox User’s Guide〞，The Math Works，Inc，1995。

[13] P.A. Nelson，S.J Elliott，〝Active Control of Sound〞，ACADEMIC PRESS，Inc，1993，p55~p59。

[14]出射忠明，〝汽車構造圖解〞，萬里機構‧萬里書店，1995，p21~p57。

[15]William H. Crouse，〝Automotive Eingine Design〞，McGraw-Hill，Inc，1994，p178~p206。

[16]胡昌華、張軍波、夏軍、張傳，〝基於MATLAB的系統分析與設計-小波分析〞，西安電子科技大學出版社，1999，p1~p5。

[17]張智星，〝MALAB程式設計與應用〞，清為科技股份有限公司，1999。

[18]葉懷仁，〝類神經網路應用於引擎聲訊之辨識〞，中山大學物理研究所碩士論文，2003。