The effectiveness of Max‐half‐Mchart over Max‐Mchart in simultaneously monitoring process mean and variability of individual observations

Abstract

Abstract Control charts are widely used in industrial environments for the simultaneous or separate monitoring of the process mean and process variability. The Max‐Mchart is a multivariate Shewhart‐type simultaneous control chart that is used when monitoring subgroups. While this sampling design allows the computation of the generalized variance (GV) used to calculate the process variability, a GV chart cannot be plotted for individual observations. Hence, we cannot compute the single statistic in the Max‐Mchart. This study aims to overcome the aforementioned issue. To this end, first, we develop a new Max‐Mchart for individual observations by utilizing the statistic in the dispersion control chart. Second, instead of the standard normal distribution, we propose a new transformation using a half‐normal distribution to calculate the statistic for the process mean and process variability. Thus, the proposed chart is called the Max‐Half‐Mchart, whose control limit is calculated using the bootstrap approach. An evaluation based on the average run length values shows the robustness of the Max‐Half‐Mchart for the simultaneous monitoring of the process mean and process variability. The single statistic in the Max‐Half‐Mchart is more consistent with the statistic in Hotelling's T2 and the dispersion chart than that of the Max‐Mchart.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 20 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.