几种检测过程方差的HWMA控制图  

Several HWMA Control Charts for Detecting Process Variances

作　　者：刘双 何川[2] 张士豪 张久军[3] Liu Shuang;He Chuan;Zhang Shihao;Zhang Jiujun(School of Mathematics and Information Science,Anshan Normal University,Anshan Liaoning 114007,China;College of Sciences,Northeastern University,Shenyang 118019,China;School of Mathematics,Liaoning University,Shenyang 110036,China)

机构地区：[1]鞍山师范学院数学与信息科学学院,辽宁鞍山114007 [2]东北大学理学院,沈阳118019 [3]辽宁大学数学院,沈阳110036

出　　处：《统计与决策》2021年第19期30-34,共5页Statistics & Decision

基　　金：辽宁省自然科学科技基金资助项目(2019ZD-0193,2020-MS-139);辽宁省科学研究经费资助项目(LSNQN201912)。

摘　　要：传统控制图包括Shewhart控制图、指数加权移动平均(EWMA)控制图、累计和(CUSUM)控制图,以及在此基础上进行修改的控制图。为了进一步提高对正态过程均值的中小漂移的检测能力,有学者提出了混合加权移动平均(HWMA)控制图。控制图可以对生产过程的质量特征进行监控,其中方差控制图是重要的研究内容之一。文章基于样本方差及其变换提出了四种检测过程方差的HWMA控制图,给出了控制图的设计方案,对控制图的性能进行了分析和比较。模拟结果表明,新提出的控制图对检测方差的中小漂移具有很好的检测能力。实例分析也验证了所提方法的有效性。Traditional control charts include Shewhart control charts,exponential weighted moving average(EWMA)control charts,cumulative sum(CUSUM)control charts and control charts modified on this basis. In order to improve the ability of detecting the small and medium drift of normal process mean,scholars proposes a hybrid weighted moving average(HWMA)control chart. Control chart can monitor the quality characteristics of production process,and variance control chart is one of the important research contents. This paper is based on sample variance and its transformation to propose four HWMA control charts for detecting process variances,then gives the design scheme of the control charts,and finally analyzes and compares the performance of the control charts. The simulation results show that the proposed control chart has good ability to detect the small and medium drift of variance. The effectiveness of the proposed method is also verified by an example.

关 键 词：方差 平均运行长度 统计过程控制 均匀加权移动平均 

分 类 号：O213[理学—概率论与数理统计]