應用熵權重法與灰色理論最佳化多品質問題

Abstract

在這競爭力激烈的二十一世紀，工業界需具備先進的技術及複雜的製程，才能因應現代消費者對產品多元化的要求，並且需不斷的創新研發產品、改善產品品質、降低生產成本才能保有市場競爭力。因此，如何在有限之實驗成本、時間及機台可行性等的情況下，有效地找到多品質製程中可控因子的最佳配方以使各種重要品質皆能被兼顧，就成為現今製造商一個非常重要的議題。工業界尋找製程最佳參數組合常使用實驗設計(Design of Experiment, DOE)，但大多只針對單品質特性進行最佳化，目前雖然已有一些關於多品質同時最佳化的相關文獻，但均有些缺失。因此，本論文之主要目的是針對利用實驗設計法規劃之具多品質特性的實驗，結合灰關聯分析法、熵權重法及雙反應曲面法，提出一套同時最佳化多個品質特性之最佳化方法。利用本研究所提出之演算法可有效找出研發新產品或改善現有製程之參數的最佳設定值。本研究最後以台灣某半導體公司之蝕刻製程為實際案例，驗證了本研究所提出之演算法確實有效可行。

Facing the sharp competitiveness in twenty-first century, the advanced technology and sophisticated manufacturing process are necessary for manufacturers to meet the consumer’s requirements. Developing innovative products, improving product quality and reducing production cost are effective ways to maintain market competitiveness. Therefore, finding the optimal factor-level combination in a multi-response process under the restricted experimental cost, experimental time and machine feasibility becomes a very important issue for manufacturers. Design of Experiments (DOE) is often applied in industry to determine the optimal parameter setting of a process. However, DOE can only be utilized to optimize single response. Although many studies have developed optimization procedures for multi-response problems, they still have some shortcoming. Therefore, the main purpose of this study is to develop a method of optimizing multiple responses simultaneously using Grey Relation Analysis (GRA), Entropy Weight Method and Dual Response Surface Methodology (DRSM). Finally, a real case from a semiconductor factory in Taiwan is utilized to verify the effectiveness of the proposed procedure.