Distribution‐free mixed cumulative sum‐exponentially weighted moving average control charts for detecting mean shifts |
| |
Authors: | Jean‐Claude Malela‐Majika Eeva Rapoo |
| |
Affiliation: | Department of Statistics, College of Science, Engineering and Technology, University of South Africa, Pretoria, South Africa |
| |
Abstract: | In this paper, we propose distribution‐free mixed cumulative sum‐exponentially weighted moving average (CUSUM‐EWMA) and exponentially weighted moving average‐cumulative sum (EWMA‐CUSUM) control charts based on the Wilcoxon rank‐sum test for detecting process mean shifts without any distributional assumption of the underlying quality process. The performances of the proposed charts are measured through the average run‐length, relative mean index, average extra quadratic loss, and average ratio of the average run‐length and performance comparison index. It is found that the proposed charts perform better than its counterparts considered in this paper under non‐normal distributions and outperform the classical mixed CUSUM‐EWMA and EWMA‐CUSUM charts in many cases under the normal distribution. The effect of the phase I sample size is also investigated on the phase II performance of the proposed charts. A numerical illustration is given to demonstrate the implementation and simplicity of the proposed charts. |
| |
Keywords: | distribution‐free control charts Monte Carlo method performance measures robust MCE and MEC schemes statistical process control Wilcoxon rank‐sum test |
|
|