Couplage de dynamiques, auto-organisation et confiance dans un système multi-agents perturbé

This paper addresses the issue of the self organization of a multi-agents system through the coupling of three dynamics : informational dynamics, behavioral dynamics and their coordination through a retro-active coupling. As an illustrative example, we consider the problem of maintaining information coherence and its robustness in a multi-agent system (MAS) that collectively gathers information from distributed sources and where some sources may be defective (deliberately or not). In this context, we propose a self-organizational approach, based on a systemic view in which we consider a structural coupling between two levels : direct information gathering at the spatial level, and communication at the social level. The local behavior rules are composed of three rule categories : rules of deployment, rules of communication, and rules of retro-action between communication and deployment. The combination of these behavior rules is then experimented in our simulations to see the emergence of roles in the system.