نوع مقاله : پژوهشی
تازه های تحقیق
عنوان مقاله English
نویسندگان English
Given the manufacturer's obligation to meet expectations and financial constraints, monitoring a large number of quality characteristics with a limited sample size poses a significant challenge for researchers in statistical quality control. According to recent research, multivariate charts based on double sampling are developed under the assumption that the number of quality characteristics is less than the sample size. The objective of this paper is to monitor the dispersion matrix of high-dimensional processes using a combination of the ridge penalized likelihood ratio statistic and double sampling method. To evaluate the performance of the proposed chart, we first introduce seven out-of-control scenarios, which include three combined diagonal/non-diagonal patterns, two diagonal patterns, and two non-diagonal patterns. We then extract the run length properties and the expected value of the sample size associated with the proposed chart through simulation experiments. The simulation results indicate that, across all three pattern categories mentioned, the proposed chart's sensitivity in detecting disturbances in the covariance matrix improves as the magnitude of shifts in its components increases. The simulation results further indicate that as the number of components influenced by the assignable cause in the covariance matrix decreases, the proposed control chart can detect process disturbances over a longer period.
کلیدواژهها English
Copyright © Zeinab Iji. Mohammad Reza Maleki. Hossein Eghbali
License
This article is released under the Creative Commons Attribution (CC BY 4.0) license. Anyone is free to copy, share, translate, and adapt this article for any purpose, whether commercial or non-commercial, as long as proper citation is given to the authors and original publication.
