Ein Framework für agentenbasierte Simulation mit räumlich expliziter Umgebung

Abstract

Agent-based simulation adopts a microscopic view on the system under consideration, representing the relevant entities in terms of agents. An agent-based simulation model thus forms a multi-agent system: a set of agents, existing and interacting in an environment. Often the environment has to be spatially explicit. So a vital requirement of a tool for agent-based simulation is to adequately facilitate modelling space. Other requirements include the specification of the agents' behaviour and interactions, and providing a suitable simulation infrastructure. The frameworkFAMOS presented in this paper extends a framework for discrete-event simulation. Major concepts comprise basing simulated space on directed graphs and modelling agent behaviours as finite-stat e machines.