Connection between micro and macro aspects of agent modeling

Abstract

This paper presents an hierarchical agent that is structured according to different types of situations it has to deal with: routine, familiar, unfamiliar situations. The perspective then shifts from the individual agent to the multiagent system. In this context, we present the idea to integrate social laws into the architecture and show how different well-known relationships among agents are supported by our architecture. In this way, we explain how to connect micro-aspects of agent modeling with macro-aspects (i.e. behavior of groups of agents). Finally, implementation and experiments demonstrated, on some scenarios of urban traffic, the applicability of ideas presented in this paper.