主成份分析法在多重品質最佳化問題上之應用－以半導體銅製程為例

Abstract

隨著產品競爭的日益激烈，多元化的品質特性需求已經成為無可避免的趨勢，因此近十幾年來陸續有學者提出同時最佳化多重品質特性的方法，然而這些方法大部份牽涉到複雜的數學模式，因此鮮少被應用於實務問題中。本研究將多變量方法中的主成份分析法(principal component analysis，PCA)與多目標決策方法中的理想解類似度順序偏好法(technique for order preference by similarity to ideal solution，TOPSIS)融入田口方法(Taguchi method)中，構建出一個簡單實用的多品質最佳化方法。此方法除了能夠找出多重品質製程的最佳參數設計外，還可以提昇工程人員對工程知識的了解。另外，本研究將利用一個半導體銅導線製程實例來說明所發展的演算流程，並將結果與田口最佳化多重品質特性的方法做一比較，以驗證本研究方法的有效性。

In recent years, optimizing multi-response problems has become an important issue. Many papers have developed various optimization techniques for multi-response problems. However, these techniques are either troublesome for determining the proper weight for each response or too complicated for engineers to use in practice. This study proposes a procedure to optimize the multi-response process based on the principal components analysis(PCA) and technique for order preference by similarity to ideal solution(TOPSIS). The optimal settings of the design parameters for the multi-response problems can be obtained easily and efficiently by the proposed procedure. In addition, engineers can advance their manufacturing knowledge through the optimizing process. The effectiveness of the proposed procedure is demonstrated by a real case provided by an integrated circuits manufacturer in Taiwan.