Authors: Luai Jaff, Cyrille Bertelle, Gerard H.E. Duchamp
Addresses: LITIS – University of Le Havre, 25 rue Ph. Lebon – BP 540, Le Havre cedex 76058, France. ' LITIS – University of Le Havre, 25 rue Ph. Lebon – BP 540, Le Havre cedex 76058, France. ' LIPN – University of Paris 13, 99, avenue Jean-Baptiste Clement, Villetaneuse 93430, France
Abstract: In this paper, we deal with some multiagent systems modelling, based on population of automata. We focus our attention with automatic computation of emerging systems. A multiscale representation is proposed here and consists in representing the internal states of an agent behaviour by a automaton with multiplicities, on the one hand and an adaptive global system behaviour by a genetic algorithm over a population of automata, on the other hand. This genetic process can lead to generate many new automata which behaviour can be eventually similar. The role played by shift operators is to identify these similar behaviours. Two applications are presented. The first one concerns adaptive strategies in game theory. The second one concerns an automatic emerging computation of self organised multiagent systems based on the efficience of operation expressivity of automata with multiplicities.
Keywords: shift operators; complex systems; complex systems modelling; self-organization; emergence; automata with multiplicities; probabilistic automata; genetic algorithms; GAs; multi-agent systems; MAS; agent-based systems; prisoner dilemma; adaptive strategies; game theory.
International Journal of Modelling, Identification and Control, 2008 Vol.3 No.1, pp.50 - 57
Available online: 10 May 2008 *Full-text access for editors Access for subscribers Purchase this article Comment on this article