It is important to investigate ways to improve the
efficiency of the statistical control charts for
attributes, because the controls for attributes typically
require very large sample sizes for quick detection of
changes in the process, and these sample sizes can be
extremely undesirable or eventually unfeasible in many
practical situations. The adaptive schemes, which vary one
or more parameters of the control charts (sample size,
sampling interval and control limit width) according to the
most recent information about the process (the position of
the last sample point in the chart), were proven to be more
efficient than the traditional (fixed parameters) scheme in
detecting small to moderate changes in the process.
Although there is a large volume of work on adaptive
schemes applied to variables control charts, there is a
lack of adaptive schemes for control charts for attributes
in the literature, despite the already mentioned need and
opportunity of investigating ways to improve the
efficiency of the statistical control charts for attributes.
This is the motivation of this work. A fully adaptive
scheme is proposed for both np and charts. The mathematical
model is developed for the measures of performance; this
model was implemented in a spreadsheet, and used for the
quantitative evaluation of the scheme efficiency, and
comparison with the fixed-parameter (traditional) scheme
and also with other adaptive schemes, with only one or two
parameters variable. The proposed scheme proved to
be significantly more efficient in the majority of the
situations of interest, answering the necessity of finding
ways to improve the efficiency of control charts for
attributes. Additional results of this work are: the
identification of the situations in which each of the
schemes here analyzed is the most efficient (or, the other
way around: indication of the most efficient scheme for
each situation), and a set of recommended values for the
design parameters for each scheme, in each situation. These
results are meant to make easier the adoption of the scheme
in practice, increasing thereby its usefulness.
|