Introduction to

Simulation and
Modeling
Azaz Ahmed Kiani

1
Guidelines
 Start working on projects/assignments from first day.

 Come prepared in the class.

 Read book (s) and Interactive Session case studies

 Remain attentive during the class.

 I will ask questions very often

Recommended books
 Modeling and Simulation, Bungartz, H.-J., Zimmer, S., Buchholz, M.,
Pflüger, D., Springer-Verlag, 2014.

 2. Simulation Modeling Handbook, A Practical Approach, Christopher A.

Chung, CRC Press, 2004.

 3. System design, modeling and simulation using Ptolemy II, Claudius

Ptolemaeus, , Ver 2.0, Creative Commons Attribution-ShareAlike 3.0
Unported, 2014

 4. Applied Simulation Modeling, Andrew F. Seila, Vlatko Ceric, Pandu

Tadikamalla, Thomson Learning Inc., 2003
Course Learning Outcome
(CLOs)
At the end of the course the students will be able to
Explain the model classification at different levels.
Analyse complex engineering systems and associated

issues (using systems thinking and modelling techniques)

Apply advanced theory-based understanding of
engineering fundamentals and specialist bodies of
knowledge in the selected discipline area to predict the
effect of engineering activities.
Analyse the simulation results of a medium sized

engineering problem.

4
Introduction to
Simulation and Modeling

Lecture: 1

5
Modeling and Simulation
 Model
 Physical, mathematical, or logical representation of a system
entity, phenomenon, or process
 Simulation
 Implementation of a model over time
 Brings a model to life
 Shows how a particular object or phenomenon will behave
 Useful for testing, analysis or training where real-world systems
or concepts can be represented by a model
 Use of models and simulations can
 Reduce the cost and risk of life cycle activities

6
Simulation Pipeline
 Simulation is the complete process of the forecasting or
replication of a certain scenario
 Such simulations are performed almost exclusively
computer-based ---- referred as
 Computer simulation
 Simulations are nothing other than
 “Virtual experiments” on the computer.
 Attractiveness of these virtual experiments is obvious:
 Astrophysics
 Study the life cycle of a galaxy
 Geophysics
 Experimental earthquakes
 Economical or military strategies

7
The Simulation Pipeline

8
Simulation Pipeline
 Modelling:
 At the very beginning we need a model, i.e.,a simplified formal
description of a suitable extract from the item of interest, which
will then serve as the basis for the subsequent computations.
 Computation:
 Model will be pre-processed (e.g., discretized) so that it is
compatible with a computer platform. The solution of this pre-
processed model requires the identification of efficient
algorithms.
 Implementation
 More generally software-development
 Computational algorithms previously determined must be
implemented efficiently (with respect to computational time and
storage complexities, etc.) on the target architecture or
architectures.
9
Simulation Pipeline
 Visualisation (data exploration):
 Data resulting from a simulation run must be interpreted
 In some Scalar quantities this will be easy
 For high dimensional data sets—extracting the relevant
information from the flood of numbers is a science of its own.
 Validation:
 How reliable are the results?
 Sources for errors lurk in the model, in the algorithm, in the code
or in the interpretation of the results.
 Compare different models, different algorithms, and different
codes, as well as simulation results with in kind experiments.
 Depending on the source of the error, the process has to be
restarted at the respective step and the pipeline has to be
traversed once more starting from this point.

10
Simulation Pipeline
 Embedding:
 Simulations take place in a context—e.g.
 Development or production process—and should be integrated
accordingly
 This requires
 Definition of interfaces
 Reasonable software engineering
 Simple testing environments, etc.

11
Introduction to Modeling
 Model is a (simplified) image of a (partial) reality
 Mathematical Modelling denotes
 Process of the formal derivation and analysis of a mathematical
model for an effect, a phenomenon, or a technical system
 Starting point is
 Informal description
 This is typically converted next into
 Semiformal description
 Finally, an additional step is required to derive a
strictly formal model
 Unambiguous, consistent — Mathematical Model
 Example
 Management of a timetable
 Room scheduling of a school
12
Models in nowadays…
 Astrophysics
 Origin and evolution of the universe
 life cycles of stars and galaxies.
 Geophysics
 Earthquakes
 Theoretical Chemistry
 Causes for certain material behaviour on the atomic level.
 Climate Research
 Global warming,
 Holes in the Ozone Layer
 Traffic Technology
 Traffic Congestion
 Traffic Routes
 Population Models
13
 City Planners
Derivation of Models
This derivation typically occurs in several steps.
STEP I
At first, one has to determine what exactly should be

modelled thereafter simulated.

 Weather
 weather in what time frame,
 the weather in what region
 in what spatial resolution
 Population growth
 Population growth in Kairo, in Egypt, or in all of Africa

14
Derivation of Models
STEP II
Next, one needs to determine which quantities play a
qualitative role and how significant is their quantitative
impact.
The optimal trajectory of the space shuttle is influenced by

the gravitation of the moon, the gravitation of Pluto and the

gravitation of this book, but not all of these are relevant for
the computation of the trajectory
One needs to determine which quantities play a qualitative role and
how significant is their quantitative impact.
Development of Dow Jones index may depend on
statements of the director of the American central bank,
statements of the authors of this book as well as the
bankruptcy of the Sultanate of Brunei
However, investors do not have to consider these
15
statements on an equal footing
Derivation of Models
This derivation typically occurs in several steps.
STEP III
Once the set of relevant quantities has been determined,
we have to direct our attention to the network of
relationships among the model parameters that have been
deemed to be important
What is a suitable instrument to best formalize the

interactions and dependencies?

 Mathematics and computer science provide a wealth of
descriptive tools and instruments

16
Example
 Algebraic equalities or inequalities to describe laws
(E =mc2)
 Systems of ordinary differential equations(differential
equations with only a single independent variable,
typically the time t), for example for the description of
growth behaviour
(y (t) = y(t))
 Graphs:
For the modelling of round trip problems (Problem of the
traveling salesman with places as vertices and paths as edges)
 Neural networks for modelling learning;
 Language concepts, to model complex software systems
 UML
 Automata and state transition diagrams 17
Analysis of Models
The analysis and assessment of models deals with the derivation of statements
in terms of their manageability and usefulness
Does a certain model have one or several solutions, or
none?
 Is there exactly one solution?
Does the solution depend continuously on the input data
(Initial values, boundary values, material parameters,
constraints, etc.).
Given this, is a solution realistic, in particular when

keeping in mind the real-time requirements?

Is the model competitive, or do there exist models with

possibly a better price-performance ratio?

How involved is the expected implementation effort?

If all these questions could be addressed satisfactorily, one could then

approach the simulation stage. 18
Classification of Models
 We will take a closer look at two:
 Discrete vs. continuous models
 Deterministic vs. stochastic models

19
Classification of Models
Discrete Models:
Exploit discrete or combinatorial descriptions

Binary or integer quantities, state transitions in graphs or

automata
Discrete models are naturally used to model discrete

phenomena
Continuous Models:
Based on real-valued or continuous descriptions

Real numbers, physical quantities, algebraic equations,

differential equations
Continuous models are employed for continuous
phenomena

20
Classification of Models
Deterministic Models:
Output of the model is fully determined by
 Parameter values
 Initial conditions.
Stochastic Models:
Models possess some inherent randomness

Same set of parameter values and initial conditions will

lead to
 Ensemble of different outputs

21
Summary

22
ANY QUESTIONS?

23