Interaction-Oriented Agent Simulations : From Theory to Implementation
Résumé
This paper deals with the software architecture for individual-centered simulations, i.e. involving many entities interacting together. Many software architectures have been developped in this context, especially many advanced - but domain specific - frameworks. Yet those frameworks imply tight software dependencies between agents, behaviors and action selection mechanisms, which leads to many difficulties in modelling and programming. We propose a method and an architecture where interactions are reified regardless of agents, in order to obtain a complete interaction-oriented design process for simulations. Then, an agent is only an entity that can perform or undergo a set of interactions, even not specifically developped for it. Thus most interactions can be re-used in many contexts. In addition, our method clearly separates knowledge about behaviors from its processing, and thus makes the design of simulations easier. Moreover, this new and user-friendly approach helps programmers to build simulations with a large number of different behaviors at the same time, especially in the context of large-scale simulations.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...