摘要
介绍了控制图的概念和3σ原理以及与控制图共生的2类错误,一是以真为假的第一类错误,一是以假为真的第二类错误。指出了控制图与假设检验一样,遵循"小概率事件不容易发生"这一统计基本原理。论述了控制图选用的显著性水平一般要比假设检验所选取的显著性水平小很多的原因。利用控制图中的落在控制限内中心线一侧打点数量这一视角,研究了位于控制限内n个样本打点中有k个落在中心线一侧相应质量系统异常判断准则的构建问题。分析出这k个点符合二项分布,给出了位于控制限内n个打点中至少有k个打点落在中心线一侧的概率计算公式。最后得到当n≤19各种样本情况的质量系统异常判断准则。
This paper introduces the concept of control chart,three principle and two types of errors from the symbiosis with control charts.The one is really false error of the first kind,the other is false is true of the second types of errors.This paper points out that control charts are the same as the hypothesis testing by following the basic statistics principle that the little probability is not prone to happen.Meanwhile,the reason why the significance level by control charts is much smaller compared to the hypothesis testing has been explained.The research has been done aiming at constructing abnormal judgment criteria of the corresponding quality system when k numbers of n sample plotting numbers from the control limit are in the central line along with the probability formula in this situation.This paper analysis that the K points accord with the two distribution and gives the probability calculation formula when k numbers of n sample plotting numbers from the control limit are in the central line.Eventually,the alarm rules of the quality system have been concluded in the case of n 19.
出处
《沈阳师范大学学报（自然科学版）》
CAS
2013年第2期242-245,共4页
Journal of Shenyang Normal University:Natural Science Edition
基金
国家自然科学基金资助项目(10471096)
关键词
质量管理
控制图
异常判断
quality management
control charts
abnormal judgment
