A COMPARISON OF SOME ROBUST BIVARIATE CONTROL CHARTS FOR INDIVIDUAL OBSERVATIONS

Abstract

This paper proposed and considered some bivariate control charts to monitor individual observations from a statistical process control. Usual control charts which use mean and variance-covariance estimators are sensitive to outliers. We consider the following robust alternatives to the classical Hoteling's T2: T2MedMAD, T2MCD, T2MVE a simulation study has been conducted to compare the performance of these control charts. Two real life data are analyzed to illustrate the application of these robust alternatives.

Keywords

• Bivariate control chart
• False Alarm
• Hotelling's T2 statistic
• Outliers
• Robust estimation
• Simulation Study
• Statistical process control