Modeling and Simulation -- Overview

People have always made models of the physical world, both to better understand it and to aid in project planning. Computers can enable very complex systems to be modeled accurately.

Modeling The Physical World

A computer model represents characteristics of some physical, social, or economic phenomenon, and attempts to reproduce or approximate its behavior and/or outputs in response to inputs

Trivial example -- the Mac’s Audio CD Player, discussed as the “perfect GUI,” attempts to replicate a physical CD player to aid users in understanding its operation

Complex example -- Boeing’s 777 was designed on-line; certain parts such as the wing used sophisticated techniques computational fluid dynamics to create an “optimal” design

Computer Modeling ...

Physical world models are used in all areas of science and engineering, they are common in finance, production, marketing and most of business

Modeling has the advantage that …

An artifact need not be created to be studied -- design

Complicated phenomena, too fast (explosive reactions), too tiny (molecular structure), too dangerous, or too distant (Jupiter) to observe, can be studied

Explanations for phenomena can be checked out before constructing an experiment or going on a field trip -- exploration

Modeling’s main disadvantage: It’s only as good as the model

Constructing A Computer Model

In a model all phenomena are represented as data (mostly numbers) and all changes are realized by computations on the numbers

To model the atmospheric pressure of an aircraft’s surface as it returns from space flight, the geometry of the plane must be represented in a computer: polygons

The Navier-Stokes equations for air flow must be applied

Simulation: Making A Model Behave

An essential tool of computer modeling is simulation

A bank can be modeled using simulated tellers and depositors and mathematics called “queuing theory”

Select a random arrival rate (frequency of new customers, and service rate -- teller speed)

Representation

A key issue in a simulation is the representation of the physical phenomena

It is not necessary to give form to all aspects of the phenomena, only those features necessary for the computation

Modeling And Simulation Accuracy

A computer model is only as good as the mathematics and programming on which it is founded

All computer models ignore features of the physical system and all make simplifying assumptions

A computer model’s predictive ability is directly related to the features ignored and the assumptions made … so, do not automatically accept a computer model any more than you would automatically accept a legal contract … Check the fine print!

Modeling And Simulation: Key Ideas

Simulations will usually have some representation of time and space

Example: in the Game of Life, a 2-D rectangular grid of cells for space, and a step count for time

Models are abstractions of the real world – we can’t represent everything, so we pick with care the attributes of interest to represent


	People have always made models of the physical world, both to better understand it and to aid in project planning. Computers can enable very complex systems to be modeled accurately.


	A computer model represents characteristics of some physical, social, or economic phenomenon, and attempts to reproduce or approximate its behavior and/or outputs in response to inputs
		Trivial example -- the Mac’s Audio CD Player, discussed as the “perfect GUI,” attempts to replicate a physical CD player to aid users in understanding its operation
		Complex example -- Boeing’s 777 was designed on-line; certain parts such as the wing used sophisticated techniques computational fluid dynamics to create an “optimal” design


	Physical world models are used in all areas of science and engineering, they are common in finance, production, marketing and most of business
	Modeling has the advantage that …
			An artifact need not be created to be studied -- design
			Complicated phenomena, too fast (explosive reactions), too tiny (molecular structure), too dangerous, or too distant (Jupiter) to observe, can be studied
			Explanations for phenomena can be checked out before constructing an experiment or going on a field trip -- exploration
	Modeling’s main disadvantage: It’s only as good as the model


	In a model all phenomena are represented as data (mostly numbers) and all changes are realized by computations on the numbers
		To model the atmospheric pressure of an aircraft’s surface as it returns from space flight, the geometry of the plane must be represented in a computer: polygons







		The Navier-Stokes equations for air flow must be applied


	An essential tool of computer modeling is simulation
	A bank can be modeled using simulated tellers and depositors and mathematics called “queuing theory”
	Select a random arrival rate (frequency of new customers, and service rate -- teller speed)


	A key issue in a simulation is the representation of the physical phenomena
	It is not necessary to give form to all aspects of the phenomena, only those features necessary for the computation


	A computer model is only as good as the mathematics and programming on which it is founded
	All computer models ignore features of the physical system and all make simplifying assumptions
	A computer model’s predictive ability is directly related to the features ignored and the assumptions made … so, do not automatically accept a computer model any more than you would automatically accept a legal contract … Check the fine print!


	Simulations will usually have some representation of time and space
		Example: in the Game of Life, a 2-D rectangular grid of cells for space, and a step count for time
	Models are abstractions of the real world – we can’t represent everything, so we pick with care the attributes of interest to represent