Floating Non-Parametric Control Charts for Process Parameter Using Computation

Abstract

When there are doubts about the intended quality attributes parametric distribution, nonparametric control charts are frequently utilized.   The majority of research in the literature on statistical process control (SPC) employed parametric techniques, wherein the functional connection remains unchanged across in-control (IC) and out-of-control (OC) scenarios. Nonparametric profiles are highly prevalent and have a distinct functional relationship under OC circumstances.   Control charts are an excellent resource for analyzing variability.   Control charts with a progressive mean and cumulative sum are useful for process monitoring.   The CUSUM chart presents data from the current and prior samples, as opposed to calculating the subgroup mean separately.   With a single sample, this graph performs well.   By adjusting the smoothing value after each sample, the progressive mean statistic is a variant of the EWMA statistic.   Every sample in the PM charts is given equal weight, taking into account both the current and all prior samples.   A floating nonparametric (FNPCUSUM) control chart was our suggestion in this situation.   A comparison was made between the proposed chart and earlier charts that were created with varying run lengths.   The effectiveness of utilizing computation to create floating nonparametric control charts is demonstrated by this comparison.   The comparison showed that, especially for minor to moderate changes in process location, the suggested control chart performed better than the other competing control charts.   Finally, to demonstrate the effectiveness of the suggested control chart, a real-world application is also provided for quality practitioners.