From above, the fit model of S01 data is a Weibull
distribution, and the maximum likelihood estimation of
the parameters is g ¼ 0:847; y ¼ 1:965: The same procedures
can be applied to all stockers. Most of the flow
move from every stocker can be fitted by the Weibull
distribution, but some cannot. From the from-to-table,
the interarrival time is not exactly an exponential
distribution. However, it can be a Weibull distribution.
An exponentialdist ribution is a specialcase of the
Weibull distribution, so the data set still can be
approximated to the same family distribution. Consequently,
the assumption of the interarrival time from the
simulation model is correct at 90% confidence interval.
Namely, the input rate is Poisson distribution and the
process flow of the interarrival time is an exponential
distribution. The zone test is also tested and divided by
etch, diffusion, thin film, photo, and implant, respectively.
The results were shown that the underlying
distributions of most stockers are exponential or of the
Weibull distribution. The values of the p-value, b; g; and
y for all stockers are shown in Table 4.