Non-MEMS精微加工的研究發展已是國際潮流,精微加工製程中微銑刀的不易準確定位卻是使此技術無法普及於工業應用上的重大瓶頸之一,因此,本論文利用機械視覺技術對於機精密刀具定位進行研究,並將此視覺系統進一步發展為微型工具機加工時的切削動態診斷系統。藉由所發展的演算法,能有效的於機上檢測出微型刀具於3-D的定位誤差,並於機上進行誤差補償的定位系統。另外,研究中也同樣以影像檢測原理發展切削動態診斷系統,其主要是希望透過對影像灰階值進行快速傅立葉轉換(FFT)後,利用灰階頻譜的主要頻率來檢測出刀具的真實刀具切削負荷(Feed/Tooth),再進一步計算兩項表面灰階值特徵係數(包含異常灰階指數與異常灰階比指數),用以判斷微型刀具切削動態穩定性。研究中,也利用實驗室的肘節式精微工具機與大立MCV-1020機台,進行所提之刀具定位與異常檢測方法進行實驗驗證,實驗結果證明該系統可精密定位微銑刀;另一系統也可成功測出刀具切削負荷,然而以表面灰階值特徵係數評估切削動態穩定性部分則仍不盡理想,有待需進一步探討之。
The development of Non-MEMS micro manufacturing technology has been the trend for modern industries. Precisely positioning the micro tool is still the bottle neck of implementation of micro milling process. In this study, a new method using machine vision technology was developed for identify the position error of a micro cutter. Based on the method, an on-machine tool positioning system that can identify the 3-D position errors of a micro tool and automatically compensate the errors for micro machining was developed. Through employing Fast Fourier Transform (FFT) and calculating two surface grey-level-value indexes, another vision-based measurement method was developed to diagnose the dynamic stability of a micro machining process. The major frequency measured from FFT was used to identity the actual chip load of the cutting. The two surface grey-level-value indexes including abnormal gray-level index and abnormal gray-level ratio were used to evaluate the smoothness of machined surface which is influenced by the stability of the cutting. Separate experiments were conducted to verify the two developed methods. The experimental results have shown that the micro cutter can be accurately positioned by the tool positioning system. The stability diagnosis system can successfully detect the chip load of micro milling. However, the capability of stability diagnosis was sill not very reliable. Further study is needed to improve the method.