基于多点报警准则的高质量过程缺陷率控制图

工业工程

基于多点报警准则的高质量过程缺陷率控制图

中原工学院系统与工业工程技术研究中心，河南郑州, 450007

出版日期:2016-08-30 发布日期:2016-10-08
作者简介:王海宇（1979-），男，山西省人，副教授，博士，主要研究方向为工业工程及质量工程。
基金资助:
国家自然科学基金资助项目（71002073）；河南省高校科技创新人才项目（15HASTIT011）

Defective Rate Control Chart for High Yield Process Based on Multi-dot Alarm Rules

Center of System & Industrial Engineering, Zhongyuan University of Technology, Zhengzhou, 450007, China

Online:2016-08-30 Published:2016-10-08

摘要/Abstract

摘要：

为了提高用于高质量过程监控的累积计点值控制图的监控效率，提出了一类采用多点报警准则的高质量过程缺陷率控制图设计，并给出了计算这类改进型控制图的平均运行长度的马尔科夫链方法。为了验证该设计方法的改进效果，分析比较了采用多点报警规则的3种累积计点值控制图（报警规则为连续两点出界、连续三点中有两点出界、连续三点出界）与报警规则不改变的（报警规则为有一个点出界）累积计点值控制图的监控效率。比较结果表明：在受控状态下平均运行长度都为370的情况下，多点报警的累积计点值控制图的失控状态下的平均运行长度分别较报警规则不变的控制图的平均运行长度减小了约44%、40%、63%，控制图发现缺陷率增大的效率明显改善。得出结论：这种多点报警的累积计点值控制图发现缺陷率增大的速度提高40%~60%。

关键词: 高质量过程, 累积计点值控制图, 平均运行长度

Abstract:

A kind of defective rate control chart for high yield process based on multidot alarm rules is proposed to improve the detecting efficiency of cumulative quantity control charts. A Markov chain method is used to calculate average run length of these charts. To illustrate the improved efficiency, the performances of three multi-dot alarm rule charts are compared with normal alarm rule chart. The results show that, when the average run length in control for the four charts reaches 370, the average run length out of control of three multi-dot alarm rule charts decreases respectively by nearly 44%, 40% and 63% of the unchanged alarm rule chart. The efficiency of detecting the increase of defective rate is improved significantly. The design method proposed can increase the speed of control chart to detect rate raise of high yield processes in 40%~60%.

Key words: high yield process, cumulative quantity control charts, average run length

王海宇. 基于多点报警准则的高质量过程缺陷率控制图[J]. 工业工程.

WANG Haiyu . Defective Rate Control Chart for High Yield Process Based on Multi-dot Alarm Rules[J]. Industrial Engineering Journal .

参考文献

［1］马林，何桢. 六西格玛管理［M］. 北京：中国人民大学出版社， 2007.
［2］GOH T N, XIE M. Statistical control of a six sigma process［J］. Quality Engineering, 2003, 15(4):587-592.
［3］CHAN L Y, XIE M, GOH T N. Cumulative quantity control charts for monitoring production processes［J］. International Journal of Production Research, 2000, 38(2):397-408.
［4］GOH T N. A control chart for very high yield processes［J］. Quality Assurance, 1987,13(1):18-22.
［5］LAI C D, XIE M, GOVINDARAJU K. Study of a markov model of a highquality dependent process［J］. Journal of Applied Statistics, 2000, 27(4):461-473.
［6］TANG L C, CHEONG W T. CCC chart with sequentially estimated parameter［J］. IIE Transactions, 2004, 36(9):841-853.
［7］TANG L C, CHEONG W T. A control scheme for high-yield correlated production under group inspection［J］. Journal of Quality Technology, 2006, 38(1):45-54.
［8］马义中，刘阳. 高质量过程控制技术的比较研究与分析［J］. 数理统计与管理. 2008, 27(3):473-479.
MA Yizhong, LIU Yang. Comparison and analysis of high quality processes monitoring technologies［J］. Application of Statistics and Management, 2008, 27(3):473-479.
［9］宋宗伟，宋向东. 高产过程的CCC-r图控制限优化［J］. 辽宁工程技术大学学报，2014, 33(4):535-538.
SONG Zongwei, SONG Xiangdong. Optimization of CCC-r chart limits of high yield processes［J］. Journal of Liaoning Technology University, 2014, 33(4):535-538.
［10］王海宇，徐济超. 基于多点报警多元统计过程控制［J］. 数理统计与管理，2006, 25(6): 709-715.
WANG Haiyu, XU Jichao. Multivariate statistical process control based on multiple dotes alarm rules［J］. Application of Statistics and Management, 2006, 25(6):709-715.