AOR Simulation (AORS)
The Entity-Relationship (ER) and Agent-Object-Relationship (AOR) modeling and simulation framework developed at the Brandenburg University of Technology, Germany, provides the XML-based simulation languages ERSL and its superset AORSL for basic and agent-based discrete event simulation. These languages allow to express the main simulation logic with high-level constructs (such as agent behavior rules) and to define special computations with programming language expressions and code snippets, for flexibility. Their use of XML greatly facilitates readability, sharing, publication and reuse of simulation models.
AOR simulations can be made user-interactive by allowing the user to "play" (that is, take control over) an agent of the simulation scenario, or by modeling the user as an agent. There are three important use cases for user-interactive simulations:
- Explorative simulation is a promising approach for teaching and training: students and trainees can learn the dynamics of a simulation model by directly interacting with it;
- Participatory simulation is a promising approach for a class of simulation problems where one (or more) of the involved human roles cannot be sufficiently faithfully modeled and therefore have to be played by human actors;
- Simulation games result from adding motivational elements to a user-interactive simulation.
AORSL also supports the modeling of cognitive agents that maintain their own beliefs about the objects in their environment (in the form of belief triples).
Agent-based simulation (ABS) is a new paradigm that has been applied to simulation problems in biology, engineering, economics and social science. In ABS, a scenario of systems that interact with each other and with their environment is modeled, and simulated, as a multiagent system. The participating agents – animals, humans, social institutions, software systems or machines – can perform actions, perceive their environment and react to changes in it. Agents also may have a mental state comprising components such as knowledge/beliefs, memories, commitments and goals.
Compared to traditional simulation methods – like mathematical equations, discrete event-simulation, cellular automata and game theory – agent-based simulation is less abstract and closer-to-reality, since it explicitly attempts to model the specific behavior of individual actors, in contrast to macro simulation techniques that are typically based on mathematical models averaging the behavior effects of individuals or of entire populations.
AOR Simulation provides an agent-based discrete event simulation framework with a high-level rule-based simulation language and an abstract simulator architecture and execution model.