摘要：In this study, we aim to build a multivariate
control chart for autocorrelated data. We use MEWMA
and MEWMV control charts which are free of normality
assumption. Time series model is then applied to tackle
autocorrelation problem in the data where the control
charts require independence assumption. The real dataset
used is the quality characteristics of white crystal sugar,
also called gula kristal putih (GKP). There are 3 quality
characteristics of GKP, namely moisture (%), color of
solution (IU), and grain type (mm). It is considered that
these quality characteristics are correlated each other. Our
results show that the variability process is out of control
where there are 5 observations outside the control limits.
Meanwhile the mean process is also out of control. The
factors causing the out of control include the workers, the
raw materials, the measurement, the machines, and the
methods. The process capability indices result in the
values less than 1 which means the process is not
sufficiently capable.
关键词：Autocorrelated data; Control
Chart; MEWMA; MEWMV;
VAR.
