In this paper, fixed control limits are shown to be an effective means of monitoring data that are autocorrelated. Autocorrelated data are encountered in continuous flow processes (i.e., chemical processes) due to the natural properties of such processes or may be the result of the use of automatic feedback control. Average run lengths (ARLs), based upon 10000 simulations of various out-of-control conditions, are presented for these control charts. Based upon the simulated ARLs, it is concluded that the fixed control limits effectively detect changes in the autocorrelated structure while maintaining the exact performance as a standard Shewhart control chart for individual measurements when the process is not disturbed. Furthermore, example data sets are plotted on these control charts to graphically present their effective and ease of use
Published in:
Engineering Management, IEEE Transactions on
(Volume:41
,
Issue:
3
)
Date of Publication: Aug 1994