Shafqat Ambreen, Zhensheng Huang, Aslam Muhammad
Department of Financial Mathematics and Statistics, Nanjing University of Science and Technology, Nanjing, 210094, China.
Department of Urology, Roswell Park Cancer Research Institute, Buffalo, NY, 14263, USA.
Sci Rep. 2023 Sep 30;13(1):16459. doi: 10.1038/s41598-023-42632-x.
Control charts are powerful tools to observe the presentation of the manufacturing process. Mostly, when the data in industries come from the process may follow non-normal or unknown distributions. So, the distribution-free control charts are useful in practice when the possibility model of the process productivity is unknown. In such situations, the correct selection of the sampling mechanism is beneficial for process examination. This paper proposes a nonparametric exponentially weighted moving average signed-rank (EWMA-SR) and also proposed a homogeneously weighted moving average Signed-Rank (HWMA-SR) control charts for examining the small shift in process with the help of an auxiliary variable (in the form of a regression estimator) by using an efficient plan, namely, a repetitive sampling plan. The proposal's presentation is evaluated and matched with its complements for different symmetric distributions by using some famous run length properties including average run length, median run length, and standard deviation of run length.
控制图是观察制造过程表现的有力工具。大多数情况下，当行业中的数据来自可能遵循非正态或未知分布的过程时，在过程生产率的概率模型未知时，无分布控制图在实践中很有用。在这种情况下，正确选择抽样机制有利于过程检验。本文提出了一种非参数指数加权移动平均符号秩（EWMA-SR）控制图，还提出了一种均匀加权移动平均符号秩（HWMA-SR）控制图，借助辅助变量（以回归估计量的形式），通过使用一种高效计划，即重复抽样计划，来检验过程中的小偏移。通过使用一些著名的运行长度特性，包括平均运行长度、中位数运行长度和运行长度的标准差，对该提议的表现进行评估，并与不同对称分布的补充情况进行匹配。
