The geometric CUSUM chart with sampling inspection for monitoring fraction defective

被引：43

作者：

Bourke, PD ^{[1
]}

机构：

[1] Univ Coll, Dept Stat, Cork, Ireland

来源：

JOURNAL OF APPLIED STATISTICS | 2001年 / 28卷 / 08期

关键词：

D O I：

10.1080/02664760120076643

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

The detection of an upward shift in the fraction defective of a repetitive process is considered using the geometric CUSUM. This CUSUM makes use of the information provided by the run-lengths of non-defective items between successive defective items, and was initially developed for the case of 100% inspection. This paper considers the geometric CUSUM under sampling inspection, and emphasizes that the pattern of sampling inspection can be quite haphazard without causing any difficulty for the operation of the CUSUM. Two separate mechanisms for the occurrence of a shift are considered. Methods for evaluating zero-state and steady-state ARL are presented for both 100% inspection and sampling inspection. Parameter choice is also considered, and recommendations made. Comparisons with some np-charts are provided.

引用

页码：951 / 972

页数：22

共 22 条

[1]

[Anonymous], APPL STAT

[2]

[Anonymous], CUMULATIVE SUM CHART

[3] DETECTING A SHIFT IN FRACTION NONCONFORMING USING RUN-LENGTH CONTROL CHARTS WITH 100-PERCENT INSPECTION [J].