Monitoring simple multivariate linear profiles in Phase II using MHWMA charts

Articles in Press

Document Type : Persian Research paper

Authors

1 Department of Industrial and Systems Engineering, Isfahan University of Technology, Isfahan, Iran

2 Department of Industrial Engineering, Yazd University, Yazd, Iran

Abstract

In many advanced industrial processes, quality can be effectively characterized using multivariate simple linear profile models, which describe regression relationships between multivariate correlated response variables and a single explanatory variable. In this study, three new control schemes based on the MHWMA control chart are proposed for Phase II monitoring of multivariate simple linear profiles. Based on a multivariate homogeneously weighted moving average control chart, the proposed schemes show enhanced sensitivity to step changes in profile model parameters. The three proposed schemes, referred to as MHWMA, MHWMA/χ², and MHWMA-2/MMECD, are designed to monitor changes in regression coefficients and the dispersion structure of the profiles through different modeling strategies. To evaluate the performance of the proposed methods, a simulation study based on the CVRL performance measure is conducted. The results indicate that the MHWMA/χ² monitoring scheme outperforms the competing methods in most scenarios in terms of change detection efficiency. The study also includes a sensitivity analysis to investigate the effects of the design parameters. Finally, to assess the practical applicability of the proposed approaches, two real-world case studies are analyzed, confirming the effectiveness of the proposed control charts under operational conditions.

Keywords

Main Subjects

Research in Production and Operations Management

Articles in Press, Accepted Manuscript
Available Online from 10 May 2026

Files

Share

How to cite

Statistics