Simulation

The simulation or simulation is an approach for analysis of systems which are too complex for theoretical or formulaic treatment. This is the case mainly in dynamic system behavior. In the simulation experiments are carried out on a model to gain insight into the real system. In the context of simulation is referred to as the simulated system and of a simulator as an implementation or realization of a simulation model. The latter represents an abstraction of the system to be simulated is (structure, function, behavior). The sequence of the simulator with concrete values ​​( parameterization) is referred to as a simulation experiment. Its results can then be interpreted and transferred to the system to be simulated.

Therefore, the first step of a simulation is always the model discovery. If a new model is developed, it is called modeling. If an existing model suitable to make statements about the problem to be solved position, only the parameters of the model must be set. The model, respectively, simulation results can then be used to draw conclusions about the problem and its solution. It can - if stochastic processes were simulated - Connect statistical analysis.

The method of simulation is used for many problems in practice. Major fields of the use of simulations, the flow, traffic, weather and climate simulation.

Subdivision

One can distinguish between with and without computer simulations. A simulation is a "Like " playthrough of processes; which you can also do without a computer.

Without a computer

Physical experiments are also referred to as simulation: A car crash test example is a simulation of a real traffic situation in which a car is involved in a traffic accident. The history of the accident, the traffic situation and the exact nature of the other party is greatly simplified. Also, no person involved in the simulated accident, crash test dummies instead be used, together with real people who have certain mechanical properties. A simulation model so only certain aspects of a real accident has in common. What aspects of this are largely depends on the question to be answered by the simulation.

Also fall into this category experiments in the flow wind tunnels. Here statements about air resistance and lift of aircraft can be made for example on a scaled-down model. The same goes for fire simulations: Hazardous situations such as fires in enclosed areas or vehicles to be adjusted and trained with real staff for training purposes of rescue and deletion or tested new materials for their fire protection properties.

Computer

If today is of " simulation " is mentioned, one thinks almost always computer simulations. Basically, the simulation can be in static vs. dynamic and stochastic vs.. deterministic simulation divided. In the static simulation time plays no role as a dynamic size, and is not part of the system. The deterministic simulation excludes random (stochastic ) events.

Reasons to use

For the use of simulations, it could be several reasons:

• A study on the real system would be too complicated, too expensive, unethical or dangerous. Examples: Driving simulator ( too dangerous in reality )
• Flight simulator for pilot training, adjustment of critical scenarios ( engine failure, emergency landing )
• Crash test ( too dangerous or too expensive to reality)
• Simulation of manufacturing systems before making any modifications (multiple retrofitting of the reality would be too complicated and too expensive)
• Simulators in medical training ( training the patient in some areas no ethical )

• Due to the system. Example: Simulation of single molecules in a liquid, Astrophysical Processes
• The real system is working too fast. Example: simulation of circuits
• The real system is too slow. Example: simulation of geological processes

Applications

From the application point of view, different simulation types can be distinguished:

Confines

Any form of simulation are also limits to which one must always be aware. The first limit follows from the limitations of the medium, ie, the finiteness of energy ( for example, computing capacity ), time and finally money. A simulation must therefore also economically make sense. Because of these limitations, a model must be as simple as possible. This in turn means that the models used often represent an oversimplification of reality. These simplifications naturally also affect the accuracy of the simulation results. The second limit follows: A model delivers only in a particular context results that can be applied to reality. In other parameter ranges, the results can be just plain wrong. Therefore, the validation of the models is an important component of simulation technology for the respective application. Possible further bounds inaccuracies in the output data (such as measurement error ) and subjective obstacles are mentioned (for example, lack of information flow on production error).