單 邊 製 程 限 制 下 選 擇 較 好 的 供 應 廠 商

Abstract

在常態製程只有單邊製程限制的情況下，製程能力指標CPU、CPL在製造業被廣泛地用來量測產品的良率。在本篇論文，我們將探討比較兩個單邊製程限制的供應廠商的問題，並選取有較高產品良率的廠商。我們發展出一個正確的方法來處理這類問題，這個方法是用假設檢定來比較兩個製程，我們導出檢定統計量的分配及機率密度函數。檢定的臨界值用來決定廠商的選取。為了業界人員的方便使用，我們將這些結果整理成表格詳列在論文末尾。最後我們利用真實的例子比較兩個WDM的供應商，介紹有了工廠資料該如何應用我們的方法去解決問題。

For stable normal processes with one-sided specification limits, capability indices CPU and CPL have been widely used in the manufacturing industry to provide numerical measures on process yield. In this paper, we consider the supplier selection problem, which deals with comparing two one-sided processes and select a better one with higher process yield. We present an exact approach to handle the supplier selection problem. A hypothesis testing for comparing two processes is considered. The exact distribution and the probability density function of a test statistic are obtained. Critical values of the test are calculated to determine the supplier selections. The results are tabulated for practitioners’ convenience. An application example on comparing two WDM manufacturing processes is presented to illustrate the practicality of our approach to actual data collected from the factories.