Microsoft Office Tutorials and References
In Depth Information
that can be useful with many simulation techniques. In cases where a commercially
available package is necessary, Excel can still have a critical role to play in the early
or rapid prototyping of problems. Rapid prototyping is a technique for quickly cre-
ating a model that need not contain the level of detail and complexity that an end-use
model requires. It can save many, many hours of later programming and modeling
effort by determining the feasibility and direction an end-use model should take.
Before we proceed, I must caution you about an important concern. Building
a Monte Carlo simulation model must be done with great care. It is very easy to
build faulty models due to careless consideration of processes. As such, the chapter
will move methodically toward the goal of constructing a useful and thoroughly
conceived simulation model. At critical points in the modeling process, we will
discuss the options that are available and why some may be better than others. There
will be numerous tables and figures that build upon one another, so I urge you to
read all material with great care. At times the reading may seem a bit tedious and
pedantic, but such is the nature of producing a high quality model—these things
cannot be rushed. Try to avoid the need to get to the punch-line too soon.
8.2 Types of Simulation and Uncertainty
The world of simulation is generally divided into two categories— continuous event
simulation and discrete event simulation . The difference in these terms is related
to how the process of simulation evolves—how results change and develop over
some dimension, usually time. For example, consider the simulation of patient
arrivals to the local hospital emergency room. The patient arrivals, which we can
consider to be events , occur sporadically and trigger other events in a discrete fash-
ion. For example, if a cardiac emergency occurs at 1:23 pm on a Saturday morning,
this might lead to the need of a defibrillator to restore a patient’s heartbeat, special-
ized personnel to operate the device, as well as a call to a physician to attend to
the patient. These circumstances require a simulation that triggers random events at
discrete points in time and we need not be concerned with tracking model behavior
when events are not occurring . The arrival of patients at the hospital is not continu-
ous over time, as might be the case for the flow of daytime traffic over a busy freeway
in Southern California. It is not unusual to have modeling phenomenon that involves
both discrete and continuous events. The importance of making a distinction relates
to the techniques that must be used to create suitable simulation models. Also, com-
mercial simulation packages are usually categorized as having either continuous,
discrete, or both modeling capabilities.
8.2.1 Incorporating Uncertain Processes in Models
Now, let us reconsider some of the issues we discussed in the Chap. 7, particularly
those in Fr. Efia’s problem of planning the events of Vegas Night at OLPS
and let us focus on the issue of uncertainty. The problem contained several
Search JabSto ::

Custom Search