双抽样均值-标准差联合控制图的构建及参数设计  

Configuration and parameters design of double sampling mean-standard deviation combined control charts

作　　者：褚崴[1] 孙树栋[1] 李彬[2] 

机构地区：[1]西北工业大学机电学院工业工程系,陕西西安710072 [2]西安交通大学管理学院,陕西西安710049

出　　处：《计算机集成制造系统》2008年第7期1281-1289,共9页Computer Integrated Manufacturing Systems

基　　金：国家863/CIMS主题资助项目(2003AA411110,2007AA04Z187)~~

摘　　要：在参考双抽样均值控制图及双抽样标准差控制图的基础上,构建了双抽样均值-标准差联合控制图。建立了控制图参数设计的优化模型,用以监控生产过程质量,并通过遗传算法完成了模型求解。采用马尔科夫链方法,完成了联合控制图平均运行长度性能指标的计算。基于该指标,通过两阶段抽样和可变抽样样本及抽样区间均值-标准差联合控制图的性能对比,分析验证了该控制图的性能优势。Based on the previous study of double sampling mean control charts and double sampling standard deviation control charts, a double sampling mean-standard deviation combined control chart was set up. The optimization model of parameter design for control chart was constructed to monitor the quality of production process, and it was solved by genetic algo- rithm. Markov chain approach was used to calculate the average run length of the control chart, moreover, in comparison to the two stage sampling combined control charts and variable sample size and sampling interval mean-range combined control charts, the double sampling mean-standard deviation combined control chart offered a better average run length performance for almost all ranges of the shifts.

关 键 词：质量监控 生产过程 双抽样均值-标准差联合控制图 参数设计 遗传算法 马尔科夫链 

分 类 号：TP272[自动化与计算机技术—检测技术与自动化装置]