基于远离中心线的打点分布设计判异准则  

Design of alarm rules based on plotting distribution away from center line

作　　者：闻学颖 孙维[1] 

机构地区：[1]大连海洋大学应用技术学院,辽宁大连116300

出　　处：《沈阳师范大学学报（自然科学版）》2014年第4期537-540,共4页Journal of Shenyang Normal University:Natural Science Edition

基　　金：辽宁省教育厅高等学校科学研究项目(20060842);辽宁省高等教育教学改革研究项目(辽教发[2009]141号)

摘　　要：介绍了生产过程稳定与异常的基本概念,指出了稳定与异常的本质区别在于质量特性服从某一分布是否出现变化和控制图上是否呈现随机分布。异常的一种形式是样本打点超出控制限,这种情况的判异可由休哈特的著名判异准则"一点出限即判为异常"来解决;另一种形式的异常是样本打点没有超出控制限,但在控制域内呈现出单调排列、连续多个样本点上下交替变动以及样本点相对于中心线波动较大等打点分布不随机的现象。对于打点分布呈现出不随机的过程异常的判断问题,则选择相对于中心线波动较大的样本打点分布,设计相应的生产过程判异准则。The article introduces basic concepts about stability and abnormality in the production process, points out the essential differences between them depend on whether certain distribution abided by quality characteristics will change, by showing on the control charts if sample plotting can arise in the form of random distribution. When quality feature obeys normal distribution, sample plotting which tends to distribute randomly approaching to the center line presents stability; otherwise abnormality. One form of the abnormality is sample dotting beyond the control limit, which can be solved by Shewhart＇s distinguished alarm rule that a point out of limit means abnormality. The other form is that in spite of sample dotting within the control limit, there exist the following dotting irrandom distributrion phenomena in the control domain such as monotone arrangement , adjacent points altermating up and down movements among many continuous sample points, sample points fluctuating greatly relative to the centre line. In view of the judging abnormality from irrandom process shown sample dotting, the corresponding alarm rules should be taken advantage of. Therefore, such sample plotting that fluctuates greatly relative to the center line should be selected and the design of alarm rules in the production can be obtained simultaneously.

关 键 词：质量管理 控制图 判异准则 设计 

分 类 号：O175[理学—数学]