THE PROCESS OF MODELING

The process of modeling is an exercise in abstraction, setting

limits, uncovering and defining relationships, and creating information
where none is available. This process can be defined by some universal
principles that are used by all modelers and that can be learned in the
classroom much more economically than they can be learned through
trial-and-error.
This course exposes some of the Principles of Modeling that have
been learned and shared by practitioners experienced in the art and
science of modeling and simulation over the last 30 years.
We will begin with an introduction of the concepts, terms, and ideas
that are essential to modeling.
We will then explain the basic philosophy and mid-set behind
modeling.
This is followed by a collection of modeling principles that have
been collected from experts.
Once you understand the principles we will describe the different
parts of a system model and go through the process of creating a model.
Finally, we will provide a complete example of a model product.
At the conclusion of this course you will have a clear picture of
what is involved in creating a model.

3.2. SYSTEM MODEL

A system model is a collection of entities that act and interact
together toward the accomplishment of some logical end (fig.1).
Output

Input
System

Fig. 1 Representation of a system

A model is an abstraction that behaves somewhat like a defined

system.
Any system can be represented at multiple levels of
abstraction.
In the real world, a system is a set of transformations that convert
input states into output states and events (fig. 2).
This general definition covers everything from the operations of a
computer to the process of driving nails.
A model is a physical, mathematical or logical representation of a
system, entity, phenomenon, or process. There is no movement in a
model.

O(t)

I(t)
MODEL

Fig. 2. A descriptive, functional, or physical representation of a system

The key is that it converts Input into Output through some defined
set of algorithms.
If the real system algorithms are defined, then they can be
captured in an abstract form as a model.
By definition the model is less exact and less all encompassing
than the original system (otherwise the model would be the system). The
model accepts a set of Input values and transforms them into Output
values.
The accuracy of the model is measured by the degree to which
model Outputs match system Outputs - given corresponding Input
values.
Webster's dictionary provides 25 definitions for the word "model",
13 as noun, 10 as verb and 2 as adjectives.
This is a good indication of generality of the term and one cause of
great misunderstunding.
We begin with the definitions that a model is:
- a noun: "a system of postulates, data, and inferences presented as a
mathematical description of an entity or state of affairs";
- a verb: to produce a representation or simulation of something;
- an adjectives: being a miniature representation of something.
From these foundations we can begin to understand how to use a
model.

How is a Model Used? A model captures the relavant information

and behavior of a system. Therefore, it can be used to study, understand,
disect, and communication information about that system.
So, a model of something is the way we execute a representation
of it. The model may be more complex, more profound, more exotic than
the thing itself, but, in relation to that thing, it is just a representation.
According with NATO MSMP 1998:
A model is a representation of a system, entity, phenomenon, or
process. Software models of specific entities are comprised of
algorithms and data.
NATO MSMP 1998

An algorithm is a prescribed set of well-defined, unambiguous rules

or processes for the solution of a problem in a finite number of steps.

NATO MSMP 1998

Data are the properties of an entity expressed in discrete

parametric values describing its attributes.

NATO MSMP 1998

Fig. 3. Examples of models

No model is a perfect representation of the system it models. A

perfect model would be an instance of the system itself. However, these
imperfections are completely normal, reasonable and acceptable. Each
model has a purpose which is a subset of the purpose or capability of the
real system it represents.
Even now maps are not true pictures of reality. Each map is a
product of compromises, omissions and interpretations. Even a good
map tells a multitude of little white lies
Mark Monmonier, Syracuse University, National Geografic, February 1998

There are twelve guiding steps to building a model of a system.

These are described above as distilled from the experiences of many
people in the field.

1. Describe It
- What is for?
- Why buid it?
2. Touch a Real One
- Size.
- Color.
- Connections.
3. Use a Real One
- Feel.
- Motion.
- Process.
4. Define the Interactions
- Input.
- Output.
- Dependencies.
5. Define State Variables
- Physical.
- Functional.
- Behavioral.
6. Define Hot Buttons
- What sells?
7. Code the Interface
- Isolate methods.
- Data.
8. Build Core Code
- Capture breadth.
- Stub details.
9. Dig Deeper
- Code details.
- Modular structure.
10. Rework
- Throw out bad ideas.
- Add detail.
11. Field Test
- Test under fire.
12. Better Next Time
- Learn lessons.
- Move on.
How Does a Model Become a Simulation? A simulation is a group
of models tied together with software and hardware that make the model

useful for some purpose. Simulations often require a standard set of

components to active the models within them.
A simulation is simply a model played out over a period of time.
Well talk more about simulations in a little bit, but right now lets talk
about models.
We saw that for a single reality we may define different models. A
model may contain models.
Our simulations are populated with entities that we call objects and
those objects are defined by algorithms and data.