Monitoring general linear profiles with between-profile correlation using the MEWMA based on U statistic

Abstract

The quality of a process can be characterized by a general linear profile. The monitoring methods for the profile process were proposed by scholars based on the assumption that observations are independent. However, in many applications, with the development of automation in industry, the process data usually appear correlation, which affects the ability of monitoring. Based on this, we develop a new multivariate exponentially weighted moving average monitoring scheme based on U statistic for a general linear profile, to further improve the performance of the existing methods. Numerical studies suggest the proposed method outperforms the existing methods under small and moderate shift sizes or moderate and strong autocorrelation. Finally, an example illustrates the implementation of the proposed method.

Keywords:

• Autocorrelation
• general linear profile
• monitoring
• MEWMA

Disclosure statement

The authors declare that they have no conflict of interest.

Additional information

Funding

The work of Dr. Feng Xu was supported by the Guangxi Natural Science Foundation Program (2021GXNSFBA220013) and the Doctoral Research Foundation of Guilin University of Technology (GUTQDJJ2020119).