本研究針對製造成本和加工品質因素,藉由田口實驗方法設計分析,以便獲得較佳的製程參數供相關業者參考。同時,本研究結合了模糊理論與田口實驗設計方法,將語意性判斷之品質特性資料量化,改善以往田口方法於整合多重品質特性無法量化之問題。研究中比較模糊語意品質特性產生之最佳參數組合,與實際量測品質特性產生之最佳參數組合的差異。並藉由灰色關聯分析(Gray Relational Analysis, GRA)與理想解類似度偏好順序評估法(Technique for Order Preference by Similarity to Ideal Solution, TOPSIS)方法進行模糊特性與量化特性之多重品質整合,並進一步進行結果之比較工作,以便驗證簡易之模糊語意判斷,是否可獲得與耗時量測所得之結果相似。由實驗結果發現此兩種整合方法,獲得之模糊特性值綜合排序結果非常相近;同樣地,獲得之量化特性值綜合排序結果亦非常接近,由此可證實模糊語意判斷在多重品質特性下參數最佳化之適用性極佳。
Based on the consideration of the manufacturing costs and production quality, the Taguchi method is adopted to achieve optimal manufacturing parameters for adornments related makers in this study. Compared with the difference between the optimal working parameter combinations obtained by fuzzy semantic quality characteristics and the actual measurements, the Gray Relational Analysis (GRA) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method are employed to integrate and compare multiple quality characteristics including fuzzy and quantitative data. According to the experimental results from the two integrated methods, it is observed that the general fuzzy and quantified ranking results of the multiple quality characteristics are very similar. The most important three affecting factors are all the same.