“…The bootstrap based‐procedure adjusts the control limits of a control chart to reach an in‐control ARL value which is at least equal to a desired value with a specific probability, say ( α )100 %. According to the previous literature, this procedure is very promising and don't have a severe adverse effect on the out‐of‐control performance of the control chart as shown in several studies (see Aly et al ., and Saleh et al ., for example). Following the same approach of Jones and Steiner and Gandy and Kvaløy's, the practitioner can use the following algorithm to adjust the control limit of the MAEWMA chart: - 1Using the available m Phase I samples, each of size n , the practitioner first calculates ; where , for the sub‐grouped MAEWMA chart and , for the individual MAEWMA chart.
- 2Assuming the true in‐control distribution is N p ( μ 0 , Σ 0 ), generate B bootstrap samples from and calculate the corresponding bootstrap estimates ; i = 1, 2,...., B ; where B is a large number, say 1000.
- 3Search for the MAEWMA control limit ; i =1,2....., B that satisfies the desired in‐control ARL; where the Phase II data are generated from
…”