COMPARISON OF THE MCUSUM CHARTS IN DETECTING THE SHIFTS OF THE PROCESS

In this paper, a comparison study of the different MCUSUM method proposed in the literature is performed. To do this, the MCUSUM methods in multivariate normal distribution are briefly introduced. In order to evaluate their performance, a comparison study is performed by simulation. In and out-of-control average run length of the MCUSUM methods in different scenarios of mean shifts, standard deviation shifts and correlation shifts have been compared. The results of the simulation study show that the MCUSUM methods have different performances in detecting the shifts of the process and based on the required application, a suitable MCUSUM chart should be selected.