田口與均勻實驗法及CAD/CAM技術分析加工參數與建立加工時間預測迴歸模型之研究
|
| 摘要 本研究旨在探討如何應用田口與均勻實驗法以及CAD/CAM技術來分析加工參數對加工時間的影響以及建立加工時間的預測迴歸模型。CAD/CAM技術在現代製造業中扮演著重要的角色，能夠自動生成加工程序，提高生產效率。而田口方法與均勻設計實驗法則是一種統計方法，能夠幫助我們規劃適當的實驗，以利建立準確的預測迴歸模型。本研究首先應用田口與均勻實驗法規劃了一系列的實驗，隨後運用CAD/CAM技術生成加工程序並量測加工時間。接著，利用多項式迴歸模型和高斯過程迴歸模型兩種方法對加工時間進行預測模型的建立。最後，比較兩種預測迴歸模型的誤差平方和大小。研究結果顯示，高斯過程迴歸模型的誤差平方和較低，顯示其預測準確度較多項式迴歸模型高。因此，若欲對加工工時進行精準預測，建議選擇高斯過程迴歸模型。
關鍵字：田口實驗、均勻實驗、CAD/CAM、迴歸模型。 ABSTRACT This study aims to explore the application of Taguchi and Uniform Design methods alongside CAD/CAM technology to analyze the impact of machining parameters on machining time and establish predictive regression models for machining time. CAD/CAM technology plays a crucial role in modern manufacturing by automating the generation of machining processes and enhancing production efficiency. Taguchi methods and Uniform Design methods, on the other hand, are statistical techniques that assist in planning appropriate experiments to facilitate the establishment of accurate predictive regression models. Initially, a series of experiments were designed using Taguchi and Uniform Design methods, followed by the generation of machining processes and measurement of machining time using CAD/CAM technology. Subsequently, polynomial regression models and Gaussian process regression models were employed to establish predictive models for machining time. Finally, the error squares of the two predictive regression models were compared. The research findings indicate that the Gaussian process regression model exhibits lower error squares, indicating higher predictive accuracy compared to the polynomial regression model. Therefore, for precise prediction of machining time, the Gaussian process regression model is recommended. KEYWORDS: Taguchi Experiments; Uniform Experiments; CAD/CAM; Regression Models. |