考慮批量間變異存在下的製程良率估計問題 - 常態漸近法

Assessing the Process Yield with Consideration of Batch Variation - Normal Approximation

廖茂原1、蔡孟萍2、陳瑩真1、許翼娟1
M. Y. Liao1, M. P. Tsai2, Y. C. Chen1 and Y. C. Hsu1

1靜宜大學 統計資訊學系
2靜宜大學 財務與計算數學系


摘要

  選擇一個良好的供應商可以明顯地降低作業成本,增加市場佔有率。品質始終是企業選擇供應商最重要的指標。在各種品質管制與品質保證活動中,製程良率指標 與良率是一對一的關係,因此其提供了一個很精確的方法衡量製程表現。本文考慮觀察值為一因子隨機效應模式下,推導出 的估計量的近似抽樣分配,並進一步提供檢定良率方法。本研究可幫助品質工程師做較可靠的決策。

關鍵字:常態近似、一因子隨機效應模式、良率。

ABSTRACT

A right supplier can obviously reduce operation cost and increase the market share. Quality is always the key factor for selecting the better supplier. Among various quality assurance activities, index provides an accurate measure for process yield because the relationship between it and the process yield is one-to-one. In this study, we derive the sampling distribution of the estimator of based on the one way random effect model. The result can help the quality engineer to make more reliable decision.

Keywords: Normal Approximation; One Way Random Effect Model; Process Yield.