No abstract

Abstract. Hydrological phenomena are often very dynamic and depend on numerous criteria. The STAFF software is an adaptive model for flood forecast based on self-organizing multiagent systems. It is operational since 2002 in the Midi-Pyrenees region in France. The aim of this paper is to show the relevance of our approach to model complex natural systems by focusing on the results, architecture and self-organization mechanisms of a real world application. The main idea is to let the artificial system self-adapt towards the adequate model by confronting it to real data, thus ensuring that the resulting model represents reality. Moreover, since the MAS is constantly adapting, we obtain a dynamic and autonomous system that can take into account any future dynamics (strong perturbations, sensor breakdowns…) and able to provide decision-makers with usable information anytime.

Combat Management System training uses simulation of an overall tactical situation. This involves the real-time management of numerous and diverse entities to keep the simulation scenario consistent in a highly dynamic environment. To address this difficult problem, we propose an adaptive multi-agent system in which each entity is considered as a smart sensor/effector mobile. The autonomy and the dynamic behaviour offered to each entity leads the simulation to self-adapt to inevitable disturbances of the user. According to the cooperation paradigm, this approach also allows the mobiles to highlights a coherent global behaviour with mutual helping. Finally, the system shows the relevance of the Emergence Technologies in the elaboration of a new generation of sensors. This software is currently under development in GATES, a project of the DCNS company. Article published by EDP Sciences c EDP Sciences, ROADEF, SMAI 2010Résumé. L'entraînement sur les Systèmes de Gestion de Combat requiert la simulation de théâtres d'opération complets. Cela implique la gestion en temps réel d'entités nombreuses et de types variés afin de conserver le scénario de la simulation cohérent malgré un environnement très dynamique. Pour aborder ce problème difficile, nous proposons un système multi-agent adaptatif dans lequel chaque entité mobile est considérée comme un capteur/effecteur intelligent. L'autonomie et le comportement dynamique apportésà chaque entité amène la simulationà s'adapter automatiquement aux inévitables perturbations induites par l'utilisateur. Suivant le paradigme de la coopération, cette approche permetégalement aux mobiles de mettre enévidence un comportement global cohérent avec aide mutuelle. Au final, le système tendà montrer la pertinence des Technologies de l'Emergence dans l'élaboration d'une nouvelle génération de capteurs. Ce logiciel est actuellement en développement dans le cadre de GATES, un projet de la société DONS.