联合监测间隔时间和幅值的自适应Max-EWMA控制图
    点此下载全文
引用本文:王煜,朱永忠.联合监测间隔时间和幅值的自适应Max-EWMA控制图[J].计算技术与自动化,2024,(1):129-136
摘要点击次数: 50
全文下载次数: 0
作者单位
王煜,朱永忠 (河海大学 理学院江苏 南京 211100) 
中文摘要:用于联合监测事件间隔时间和幅值的控制图多数假定数据的漂移大小已知,但在实际生产中由于数据的漂移大小未知,这些针对特定漂移大小的控制图存在监测性能下降的问题。为此,基于事件间隔时间和幅值漂移程度的无偏估计量,构造了EWMA统计量的自适应平滑参数,通过结合Max-EWMA统计量的结构特征,提出了一种新的联合监测间隔时间和幅值的自适应Max-EWMA控制图(称为Max-AEWMA图)。模拟仿真表明:在监测间隔时间和幅值的不同程度的漂移时,本文提出的控制图一致优于已有的Max-EWMA控制图,例如,a=4,b=1,λE=0.01,δa=0.59,δλE=0.73时,Max-AEWMA的ARL1=4.94,Max-EWMA控制图的ARL1=9.68。最后还通过实证分析验证了本文方法的优越性。
中文关键词:自适应  Max-EWMA  幅值  统计过程控制
 
An Adaptive Max-EWMA Control Chart for Simultaneous Monitoring of Internal Time and Amplitude
Abstract:Most control charts used to jointly monitor time between event and amplitude assume that the process shift size is known, but in actual production these control charts for specific shift sizes suffer from monitoring performance degradation because the shift size of the data is unknown. To this end, this paper constructs adaptive smoothing parameters for the EWMA statistic based on unbiased estimates of time between event and amplitude, and proposes a new adaptive Max-EWMA control chart for joint monitoring time between event and amplitude (referred to as Max-AEWMA chart) by combining the structural features of the Max-EWMA statistic. Simulations show that the proposed control chart is consistently better than the existing Max-EWMA control chart when monitoring different size of shift in time between event and amplitude. For example, when a=4,b=1,λE=0.01,δa=0.59,δλE=0.73, Max-AEWMA’sARL1=4.94, Max-EWMA’sARL1=4.94. The paper also concludes with an empirical analysis to verify the superiority of the method.
keywords:adaptive  Max-EWMA  magnitude  statistical process control
查看全文   查看/发表评论   下载pdf阅读器