本文利用統計軟體Minitab以實驗設計方法探討取樣方法對於成形螺孔之牙型填充率及成形最大扭矩值優化結果的影響，為了降低其他因素對結果的干擾，本文亦設計及提供客觀的數據取樣方法。在攻牙扭矩部分，本文選擇一組實驗、九組實驗和標準差最小的四次實驗作為取樣手法；牙型填充率部分則選用螺孔完整四牙、左側二牙和右側二牙作為取樣手法。根據上述取樣方法，我們建立了與其對應的迴歸模型並進行殘差分析的檢測以確保迴歸符合迴歸理論基本假設。後續藉由檢驗點誤差分析和模方分析研究各取樣方法之間之差異性。最後，本文將對各迴歸模型進行參數優化，以限制最佳化問題來評估取樣方法可能會造成優化分析結果的影響。透過上述分析結果，期望能供業界做為探討螺孔加工製程研究的參考，從而優化和改進現有的製程技術。;This study utilizes the statistical software Minitab and experimental design methods to investigate the impact of sampling methods on the optimization results of filling rate and maximum forming torque in threaded hole formation. To minimize the interference from other factors, this study also designs and provides objective data sampling methods. For the torque analysis, a single experiment, nine experiments, and the four experiments with the smallest standard deviation are selected as sampling techniques. For the filling rate analysis, complete four teeth, left tooth, and right tooth are chosen as the sampling methods. Based on the aforementioned sampling methods, corresponding regression models are established and residual analysis is conducted to ensure the regression meets the basic assumptions of the theory. Subsequently, point error analysis and model variance analysis are employed to examine the differences among the sampling methods. Finally, parameter optimization is applied to each regression model to evaluate the potential impact of the sampling methods on the optimization analysis results by limiting the optimization problem. Through the analysis results, it is expected that this study can serve as a reference for the industry in exploring threaded hole processing research, thereby optimizing and improving existing manufacturing techniques.
