Evacuation Process Simulation - Classification of Models

Classification of Models

Simulations are based on mathematical models that resemble reality. The calculations are usually carried out using computers. Models are the specification of a certain theory. They can be classified according to the following criteria:

(Gershenfeld, 1999):

specific -- general
phenomenological -- first principles
discrete -- continuous
numeric -- analytic
stochastic -- deterministic
quantitative -- qualitative
macroscopic -- microscopic

Simulations are not primarily methods for optimization. To optimize the geometry of a building or the procedure with respect to evacuation time, a target function has to be specified and minimized. Accordingly, one or several variables must be identified which are subject to variation.

Queueing models belong to the macroscopic models which are based on the graphical representation of the geometry. The movement of the persons is represented as flow on this graph. Microscopic models on the other hand are based on a detailed representation of geometry and population. If the individuals are autonomous and interact with each other, these models are called Multi-agent systems. Stochastic parameters describe the agents' movement and decisions and represent influences not further specified or which cannot be quantified directly and have to be calibrated via comparison with empirical data. Analytic results are very hard to obtain for social systems. General models can be applied to the evacuation of buildings, aircraft, and ships alike.

Read more about this topic:  Evacuation Process Simulation

Famous quotes containing the word models:

    Friends broaden our horizons. They serve as new models with whom we can identify. They allow us to be ourselves—and accept us that way. They enhance our self-esteem because they think we’re okay, because we matter to them. And because they matter to us—for various reasons, at various levels of intensity—they enrich the quality of our emotional life.
    Judith Viorst (20th century)