MODELING SELF-ORGANIZING SYSTEMS WITH SOCIAL INSECTS ALGORITHMS

Cyrille Bertelle (1) , Antoine Dutot (1) , Rawan Ghnemat (2)

1 RI2C - LITIS - Equipe Réseaux d'interactions et Intelligence Collective
2 GJU - German Jordanian University

Self-organization is common in natural systems. This tutorial describes some of these systems, specifically from insect societies like in bees, termites or ant colonies. In a first part, a modeling process is explained. Objects and phenomena targeted by these methods are presented. Natural or social complex systems are the context of these objects and phenomena. Basic algorithms presented for example in [8] are given. These algorithms belongs to the class of swarm intelligence methods describing how a network of interacting entities can lead to emergent properties of the whole system. In a second part, more original applications are presented, based on extensions of these basic algorithms in order to model ecosystems, urban dynamics or to propose a decentralized method to distribute simulations over dynamical communication graphs.

Mots clés: swarm intelligence; complex systems; ant colony optimization; self-organization; emergent properties

Domaines: Intelligence artificielle [cs.AI]

Cyrille Bertelle, Antoine Dutot, Rawan Ghnemat. MODELING SELF-ORGANIZING SYSTEMS WITH SOCIAL INSECTS ALGORITHMS. Brain - A Journal of Neurology , 2010, 1, pp.18-32. ⟨hal-03313596⟩