Applying simulation requires estimating the probability distribution
of interarrival times and of service times for each of the two classes
of customers. |
The pattern of interarrival times is different for merchants and other
customers. Because the downtown area is small, most of the merchants walk
to the bank and arrive randomly (i.e., an exponential distribution for
interarrival times). As described in the first simulation demo, other customers
normally park in the bank parking lot and have their parking validated
before entering the bank, which leads to a translated exponential distribution
for interarrival times. |
In comparison to other customers, a merchant's business in the bank
tends to be both more extensive and better organized in advance. Therefore,
the service time required by a merchant tends to be a little longer, but
more predictable in falling within a certain range. |