Modeling, what is:
Model a system means describe its features and behavior in terms of
components exchanging messages. In detail, each component is made of
states and activities that allow transitions between states.
It is not necessary to start with a completely detailed model: you can begin with a simple
model of the system and gradually refine it adding new details as you validate
the existing and implement any changes or improvement.
The step-by-step technique enables you to achieve good
approximations of very complex systems very quickly and the model becomes more and more
similar to the real system at any step.
Simulation, what is and what for:
Simulate a model means analyze the time evolution of the model
behavior, examining the message flow, the duration of activities, the
frequencies of states. Parameters may be set in the model, so that
simulation allows analyze different scenarios of performance, in order
to highlight critical points. In other words simulation allows to investigate different hypotheses at reduced
costs (What if analysis).
Simulation provides a method for checking your level of understanding of your
system and to produce better results
faster.
Simulation is an efficient communication tool, to show the system behavior and
to understand how it can be improved.
Simulation, why:
Simulation involves designing a model of a system and carrying out
experiments on it as it progresses through time.
Simulating a systems is very important because you can:
- Predict the course and the results of any actions
- Understand why events occur
- Identify problems and bottle-necks before implementation
- Identify critical aspects like inefficiencies
- Explore the effects of modifications
- Verify that all variables and parameters are known and used and all the
system is tested
Modeling and Simulation, how to face it:
Modeling is a powerful and communication tool.
You use models to investigate systems which are too complex to
be analyzed via mathematical or analytical functions or flowcharts.
You use models to test different hypotheses at a reduced cost.
Modeling shows how a system works and helps you
to refine and to improve it.
Models make design cycles faster, reduce costs, and enhance knowledge.
Simulation involves designing a model of the system and making
experiments on it, simulation runs.
The goal of simulation runs is to determine
how the system behave and to observe the effect of changes to the system as
events occur.
Static model
Static model describes the system mathematically, in terms of equations.
Static model does not consider timing. You can not use it to
determine the impact of event in relation to other events.
Static model does not consider the correlation of the system components, actions
associated to different events can have a different effects on
the system than their individual effects would indicate.
Dynamic model
Dynamic model (also known as simulation) is the representation of
the dynamic or time-based behavior of a system.
Dynamic model takes into account the events occurrences and can consider the
effects of variances or randomness.
Discrete event simulation
Discrete event simulation permits to model system state changing as events
occurs, like in FSM (Finite State Machine) and to consider timing
Modeling language
To correctly model your system, you need a formal, graphical and easy to use
language, with few symbols and high expressiveness capacity.
The modeling language allows to represent events, queues, actions, priorities,
policies, timing and data, and it must comprise object oriented methodology.
Suitable graphical software tool
To create the model of the system to validate it and to simulate it you need
a complete suite of graphical software tools.
The model should be easily integrated with already existing libraries and user
code.
Besides modeling you need a powerful environment to validate and debug the
simulated model, having the capability to investigate system status and values
at any time, and to change any condition on the fly, to complete debug and
investigate states, examine queues, values and events.
Faber Solution
A clear and detailed model allows an accurate analysis of the system
with a short return on investment. With Faber the model is highly maintainable and can be
rapidly adapted to the changes introduced in the real system. In this way
the model will represent faithfully the real system behavior and will be
ready for the next phase of analysis and improvement.
The Faber solution is all of these, and
more...
|
