Meningkatkan Prestasi Carta Kawalan Reja Untuk Mengawal Data Berautokorelasi
Loading...
Date
2013-05
Authors
Sie Ling, Lau
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The autocorrelation problem exists m most manufacturing industries today. To
overcome the autocorrelation problem, numerous methods have been suggested by the
relevant authors such as Montgomery and Mastrangelo (1991), Runger and Willemain
(1995), Atienza, Tang and Ang (2002) and other authors. A common method that has
been suggested is to use a two-step method which involves the use of a time series
modeL This two-step method is also called the 'model-based' approach. One of the
control charts that is constructed from a 'model-based' approach is the residual control
chart. However, a residual control chart is not sensitive to small shifts in the process
mean. This study has suggested an easy approach to improve the performance of a
residual control chart by applying the 2-of-2 runs rule and 2-of-3 runs rule on the
residual chart By using these runs rules, it is found that the performance of a residual
chart has improved, i.e., it can detect small shifts in the process mean quickly. Although
in this dissertation, we only consider an AR( 1) time series model as a research example,
it should be noted that the use of runs rules can be extended to other time series models.
Description
Keywords
Mathematics , General