Zhang extends the traditional EWMA control chart to the case of stationary processes, showing how the control limits can be analytically adjusted as a function of the process variance and autocorrelation. When parameters are unknown, the author also suggests how to estimate the control limits. Zhang's proposal is operationally simple and fully automatic. For this reason, it looks quite appealing from the practitioner's point of view. However, the impact of parameter estimation on the chart performance is not investigated. Since charts designed using estimated parameters can have an unsatisfactory behaviour, we study the impact of the estimation errors and give practical recommendations on the smallest sample size for which satisfactory accuracy is obtained.
The EWMAST Control Charts with Estimated Limits: Properties and Recommendations.
CAPIZZI, GIOVANNA;MASAROTTO, GUIDO
2007
Abstract
Zhang extends the traditional EWMA control chart to the case of stationary processes, showing how the control limits can be analytically adjusted as a function of the process variance and autocorrelation. When parameters are unknown, the author also suggests how to estimate the control limits. Zhang's proposal is operationally simple and fully automatic. For this reason, it looks quite appealing from the practitioner's point of view. However, the impact of parameter estimation on the chart performance is not investigated. Since charts designed using estimated parameters can have an unsatisfactory behaviour, we study the impact of the estimation errors and give practical recommendations on the smallest sample size for which satisfactory accuracy is obtained.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.