Abstract
In modern quality control, high-dimensional process control has become increasingly important. In addition to multivariate control chart, sampling strategy in context of limited resources with only partial information available was developed recently. In this article, we propose the sampling strategy which incorporates two new methods. First, we improve statistics in order to choose observing control characteristics in consideration of the correlations. Second, we separate the statistics for selecting characteristics and statistics for determining whether they are in-control or out of control. The two simulations show that this new sampling method has higher detection ability when the
mean shift occurs than the conventional method which does not consider the correlations especially in the case that the number of control characteristics is much bigger than that of characteristics which can observed .
