Microsimulation

Most things around us and situations are composed of several individual components. If the decomposed in a simulation examinee into its components, and uses the simulation models of these components, and not a model of the existing from them whole, are, for example, individual gas atoms to simulate the movement of a gas cloud or individual vehicles for the simulation of the traffic flow simulation, one speaks of a " microscopic simulation " and borders such models thus on a macro simulation.

Micro-simulation in economics and the social sciences

Here micro- simulations are used in various places, which may be under the same term to very different types of simulations.

Heuristic microsimulation

Or ... strategic simulation: It is used mainly in management. Mostly it involves macro simulations within the meaning of cybernetic control loops, often in proximity of simulation games, but it can also come micro simulations are used. In the physics of traffic heuristic micro simulations are used intensively, here is the simulated system consists of individual road users. Characteristic of the heuristic simulation is the most widely executed starting missing link to data and the emphasis on interaction effects between micro units. In sociology, such micro- simulations have been used since the 1970s. Have become known computer tournaments by Robert Axelrod in the field of game theory research, but in addition are also methods of statistical physics, and especially today, computer science applied ( multi-agent systems (MAS ) and multi-agent simulation).

Economic microsimulation

In the field of empirical economic research in the 1950s was the area of ​​the " Policy Evaluation Simulation ", ie simulations that attempt based on statistical data to assess the impact of policies. Since it often comes to distribution effects and non- linear changes here, macro simulations are often unsuitable here. The simplest method of statistical micro-simulation, the "static aging" established: one formulated the guidelines by the policy measure by macro time series, then selects person weights so that the requirements have been met and reads from the effects of other micro variables. Slowly only since the 1990s, come on methods of dynamic microsimulation. Here individual variables are set in quantitative relationships and these are statistically estimated. Example: The probability of becoming unemployed is a regression function of education, industry, professional position or income. In the simulation are then generated by the so- estimated models for other boundary conditions or time ranges data.

Economic micro simulations are based on representative statistical data, such as the micro-census, the SOEP or the EVS. They contain, in contrast to heuristic microsimulation little interaction effects (possibly between members of a household ).

Economic micro simulations are extensively used in recent times by the Federal Ministries ( Finance, Economics, Employment, Health ) for their planning.

Methods

To implement a micro-simulation are a number of methods: cellular automata have their origin in mathematics or computer science. The transition from cellular to discrete multi-agent simulation is sometimes flowing, sometimes simply the name of the field -dependent. The finite element method is used extensively in the engineering sciences.

Other examples

FHP model, evacuation simulation, Nagel-Schreckenberg model, VISSIM, lattice gauge theory

• Applied computer science
• Empirical Research in Economics