利用R語言以複式模擬法構建製程良率之信賴區間

Construct the Confidence Interval for Process Yield Using Bootstrap Simulation with R

蘇明鴻、陳德正、顏鏡峰
M. H. Shu, D. Z. Chen and J. F. Yan

國立高雄科技大學 工業工程與管理系


摘要

競爭激烈與迅速變遷的企業環境內,廠商為了在合格訂單中勝出,就必須了解消費者的需求,而在生活品質快速提升的大環境裡, 消費者對產品的需求逐漸由成本的依賴轉為品質的重視。因此,近年來 為了提升產品的品質提出多品質製程能力的觀念將多項品質特性的產品進行監控。
製程能力分析(Process Capability Analysis)是 SPC 中重要技術之一,目的是藉由量化的製程能力指標(Process Capability Index, PCI)來評估製程實際績效的優劣,業界常用製程能力指標有等。
製程良率(process yield)與製程能力指標有密不可分的關係,製程良率除了提供給品管人員來判斷製程或產品的優劣之外,也是公司對外展示產品良好的依據。當下有指標和指標和製程良率有一對一直接關係,指標必須使用在製程平均數等於規格中心線,然而指標並無規範,所以使用起來更廣泛。業界常用的指標和製程良率只有不等式關係。

本研究之主要目的是藉由指標與製程良率的一對一關係,算出合理的製程良率的方法,藉由重複抽樣的複式模擬法(Bootstrap Simulation)來模擬出製程良率的信賴區間。

關鍵字:製程能力指標、指標、信賴區間、複式模擬法。

ABSTRACT

In an enterprise environment where fierce competition and rapidly changing, in order to win in qualifying orders, manufacturers must understand the needs of consumers. In the environment of rapid improvement in the quality of life, consumers’ demand for products gradually shifts from cost dependence to Quality is valued. Therefore, in recent years, the concept of multi-quality process capability has been proposed to improve the quality of products.
Process Capability Analysis is one of the important technologies in SPC. To assess the pros and cons of the actual performance of the process by Process Capability Index, The commonly used Process Capability Index include .
Process yield and Process Capability Index are inseparable. In addition to providing quality control personnel to judge the quality of the process or product, process yield is also a good basis for the company to display products. There is a one-to-one direct relationship between the  Index and the  Index and the process yield. The  Index must be used in the process average equal to the specification centerline. However, the  Index is not standardized, so it is more widely used. The commonly used Index and process yield in the industry are only inequality.

The objective of this study is to use a bootstrap simulation to construct the bootstrap con_dence interval for the process yield based on Spk.

Keywords: Process Capability Index; Index; Confidence Interval; Bootstrap Simulation