Application of the belief function theory to validate multi-agent based simulations

Alexandre Veremme (1) , Eric Lefevre (2) , Gildas Morvan (2) , Daniel Jolly (2)

1 HEI - Hautes Etudes d’Ingénieur [Lille]
2 LGI2A - Laboratoire de Génie Informatique et d'Automatique de l'Artois

In this article, an architecture to validate agent based simulations is presented. The proposed pyramidal architecture is based on the belief function theory to represent and handle imperfect information at three different levels. We are then able to provide, at many steps during the simulation, the validity state compared to the real system. First results of this architecture are presented within the framework of an application in forensic entomology.

Mots clés: belief functions; multi-agent based simulation; validation

Domaines: Système multi-agents [cs.MA]; Modélisation et simulation

Alexandre Veremme, Eric Lefevre, Gildas Morvan, Daniel Jolly. Application of the belief function theory to validate multi-agent based simulations. First int. workshop on the theory of belief functions, Apr 2010, Brest, France. ⟨hal-03356620⟩