A multivariate synthetic control chart for monitoring covariance matrix based on conditional entropy

Abstract

In multivariate statistical process control field, besides monitoring the changes in the mean vector of a multivariate process, it is important to detect the changes in the covariance matrix of a multivariate process. This paper proposes a multivariate synthetic control chart for monitoring the changes in the covariance matrix of a multivariate process under multivariate normal distribution. The proposed control chart is a combination of the traditional control chart based on conditional entropy and the conforming run length chart. The operation and design of this control chart are described. © 2013 Springer-Verlag Berlin Heidelberg.