以灰色理論結合聲波訊號之微鑽孔孔徑及刀具壽命即時預測系統

Development of a Real-time Micro Drilling Diameters and Tool Life Forecasting System Based on Grey Theory and Acoustic Signal

陳峻宇(Chun-Yu Chen)

指導教授 ： 黃博滄

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2016年)

https://doi.org/10.6840/cycu201600737

摘要

隨著近年消費趨勢、加工材料與工法的改變，加工產業已從傳統的大量生產轉向輕薄短小的高品質精密加工，使得生產品質的控管與預測越來越受到重視，更成為近年來各界研究探討建立預測系統的重點，對於品質預測系統的建立方法有傳統的統計預測、模擬建模、回歸分析及個類軟性計算等方式，更有研究加入接觸式的力感測器、聲射感測器與加速規協助數據收集與投入預測補正，且都保有一定的準確率。但是上述所提及的預測模型皆需要大量數據與時間進行建模，加工參數條件一旦改變，就需要重新建模或調整系統，對於目前高精密少量多樣的加工趨勢相當不便；而且加工中的切削液與碎屑對於加裝的接觸式感測器都會形成容易故障的環境。故本研究將採用灰色理論結合非接觸式的聲波訊號感測器，希望透過少數據建模與聲波訊號補正，針對微型鑽孔加工建構一個不受加工參數條件影響的孔徑與刀具壽命即時預測系統。
　　本研究針對微鑽孔孔徑與刀具壽命所開發之灰色理論結合聲波訊號預測系統，利用灰色預測小樣本建模的特性為基礎，透過灰生成，將孔徑及聲波訊號數據輸入並顯著化其趨勢，分別預測出孔徑及聲波訊號值。再透過線性趨勢與線性插值法將聲波訊號補正回孔徑預測值，去預測孔徑與刀具壽命。本研究所使用的方法在各種加工參數與環境下，皆可達到快速建模、即時預測的目標。
　　為證實所提出的方法其可行性與準確性，本研究將設置兩組不同加工參數，進行微鑽孔加工，透過金相式電子顯微鏡與聲波訊號感測器，將孔徑與聲波訊號值投入灰預測系統，再將聲波訊號預測值補正回孔徑預測值，藉此得到更準確的各孔孔徑預測與整體刀具壽命，並探討其準確性。孔徑部分兩組補正值皆比原始預測精準且顯著，刀具壽命部分補正後平均精準度高達96.97%，比原始預測87.795%提升將近10%的準確率，故得以證實本研究所提出之預測系統其可行性與準確性。

關鍵字

灰色預測 ； 灰色生成 ； 聲波訊號 ； 微型鑽孔 ； 線性趨勢

並列摘要

As the consumer trends and the change of processing in recent years, production orientation of machining process has evolved from mass production to high quality, small-scale production; therefore, quality control and predict is becoming increasingly important. Various prediction methods such as tradition statistical forecasting, simulation modeling, regression analysis and soft computing, have been proposed to construct prediction systems in these years, and some studies even added contact force sensor, acoustic emission sensor and accelerometers to assist data collection and to increase accuracy. However, these prediction models need huge data and time to create, if processing parameters changed, it has to reboot system, which is unfavorable to high-precision machining and small-scale productions. Cutting fluid and chips are also have bad effect to contact sensors. Due to the inconvenience, in this research will implement an in-process micro drilling diameters and tool life forecasting system which integrate grey theory and non-contact acoustic sensor to correct and without the influence of the changed process setting.
　　The grey theory integrate acoustic signal in micro drilling diameters and tool life forecasting system proposed in this research is develop through grey theory and acoustic signal sensor, a method which emphasizes the input data trend and uses its small sample feature as the fundamental structure for building a real time prediction system. In this research using grey theory to forecasting cutting diameters and acoustic signal, then correction acoustic signal to diameters by linear interpolation. The grey and acoustic signal prediction system can quickly predict micro drilling diameters and tool’s operating life under specific processing settings without considering the type of tool and work part being used, the machining parameters and environment.
　　In order to prove the proposed method is both accurate and reliable, two different sets of machining parameters are used to perform micro drilling. First, an optical instrument is used to measure the diameter of the machined holes and acoustic signal sensor is used to measure the noise .Then using a small amount of the measured data are placed in the grey prediction system for the forecasting of hole sizes and tool life. Based on the information obtained, we can use hypothesis testing and (Mean Absolute Percentage Error , MAPE) test to determine a tool’s operating life and investigate its accuracy to verify the feasibility of the prediction system proposed in this research.

並列關鍵字

Grey prediction ； Grey generation ； Acoustic signal ； Micro drilling ； Linear trend

參考文獻

陳清豪, & 洪志真. (2009). 監控線性趨勢製程之研究.

Abu-Mahfouz, I. (2003). Drilling wear detection and classification using vibration signals and artificial neural network. International Journal of Machine Tools and Manufacture, 43(7), 707-720.

Ahilan, C., Kumanan, S., Sivakumaran, N., & Dhas, J. E. R. (2013). Modeling and prediction of machining quality in CNC turning process using intelligent hybrid decision making tools. Applied Soft Computing, 13(3), 1543-1551.

Bhuiyan, M. S. H., Choudhury, I. A., Dahari, M., Nukman, Y., & Dawal, S. Z. (2016). Application of acoustic emission sensor to investigate the frequency of tool wear and plastic deformation in tool condition monitoring. Measurement, 92, 208-217.

Bustillo, A., & Correa, M. (2012). Using artificial intelligence to predict surface roughness in deep drilling of steel components. Journal of Intelligent Manufacturing, 23(5), 1893-1902.