計數值允收抽樣計畫與作業特徵曲線之設計－圖形化介面程式設計使用LabVIEW語言

Design of an Attribute Acceptance Sampling Plan and its Operating Characteristic Curve: a Visual Approach using LabVIEW Programming Language

張珩(H. Chang) ； 黃聰耀(T.Y. Huang)

《華岡工程學報》 19期 (2005/06) Pp. 83-91

https://doi.org/10.29987/CCUHKJE.200506.0008

摘要

計數值允收抽樣檢驗為一最常使用的抽樣類型，本文中分成三部分將抽樣檢驗之方法程式化，第一部分藉由作業特徵曲線之設計與製作，可觀察與比較當允收機率使用二次分配機率時真確計算與卜氏分配的近似計算差異，進而在改變允收參數：n、c值時觀察作業特徵曲線的變化；本文中的第二部分則引導進入抽樣檢驗的設計方案，由單獨設定允收參數的允收水準(AQL)與生產者風險(α)，該作業特徵曲線可以圖形化呈現比較其差異性，亦可藉由單獨設定允收參數的品質界線(LQL)與消費者風險(β)，設計滿足該條件之允收抽樣計畫，且圖形化作業特性曲線可顯示不同計畫之差別；第三部分中則以同時設定計畫參數的允收水準(AQL)與生產者風險(α)及品質界線(LQL)與消費者風險(β)，以尋求能滿足該設計之抽樣計畫，並圖形化最近似該條件之多個計畫作業特徵曲線，以供使用者選擇應用。

關鍵字

計數值抽樣檢驗 ； 作業特徵曲線

並列摘要

In this article, a very useful software tool is created for designing attribute sampling plans. Three concepts associated with the design of attribute acceptance sampling plans are illustrated by using the LabVIEW programming language. The first concept is to calculate the probability of lot acceptance, based on which the operating characteristic (OC) curve of an attribute sampling plan is drawn. The two-type OC curves (type A and type B) based on the hypergeometric and binomial distributions can be shown visually by setting values of sample size (n) and acceptance number (c). The exact probability of lot acceptance can also be approximated by the Poisson distribution for large lot. The approximated probability of lot acceptance is shown interactively by changing either one of the two parameters n and c of the sampling plans. The second concept introduces the design of an attribute sampling plan based on the Acceptable Quality Level (AQL) associated with Producer's Risk (α) or the Limiting Quality Level (LQL) with Consumer's Risk (β). Several plans may all satisfy the criteria and can be visually drawn for comparison. The third concept links to design the sampling plan based on both the Acceptable Quality Level (AQL) and the Limiting Quality Level (LQL). Because the criteria are more stringent, it can be difficult to find a sampling plan that exactly satisfies both the producer's and consumer's stipulation. A preference to satisfy exactly either the producer's or consumer's risk is then incorporated in the design procedure. The decision-making framework based on these stipulations is visually displayed to aid in selecting the appropriate sampling plan.

並列關鍵字

Attribute acceptance sampling ； Operating characteristic curve