Sensitised multivariate homogeneously weighted moving average charts for Phase II monitoring of the multivariate process mean

In some processes, the occurrence of assignable cause results in small disturbances in the process parameters. Moreover, a fundamental assumption in designing control charts is that the measurements are accurate. This article aims to enhance the multivariate homogeneously weighted moving average (MHWMA) chart with 2-of-2 and 2-of-3 sensitisation rules for the rapid detection of mean changes. Then, the proposed charts are developed based on additive covariate model to account for the impact of measurement errors. Finally, multiple measurement technique is utilised to reduce the error impact on run length of the sensitised charts. The performance of the proposed charts is assessed through extensive simulations. The results show that the MHWMA chart based on the 2-of-3 rule outperforms its counterpart which employs 2-of-2 rule. Secondly, the measurement error negatively affects the power of both the proposed charts. Thirdly, increasing the number of measurements for each product can improve the chart detectability.