Communication Dans Un Congrès Année : 2014

A Distributed Decision-Theoretic Model for Multiagent Active Information Gathering

Jennifer Renoux
  • Fonction : Auteur
  • PersonId : 962475
Simon Le Gloannec
  • Fonction : Auteur
  • PersonId : 830648

Résumé

Multirobot systems have made tremendous progress in ex-ploration and surveillance. In that kind of problem, agents are not required to perform a given task but should gather as much information as possible. However, information gather-ing tasks usually remain passive. In this paper, we present a multirobot model for active information gathering. In this model, robots explore, assess the relevance, update their be-liefs and communicate the appropriate information to rel-evant robots. To do so, we propose a distributed decision process where a robot maintains a belief matrix representing its beliefs and beliefs about the beliefs of the other robots. This decision process uses entropy and Kullback-Leibler in a reward function to access the relevance of their beliefs and the divergence with each other. This model allows the derivation of a policy for gathering information to make the entropy low and a communication policy to reduce the di-vergence. An experimental scenario has been developed for an indoor information gathering mission.

Fichier principal
Vignette du fichier
activePeceptionAAMAS14.pdf (393.45 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence
Loading...

Dates et versions

hal-01096027 , version 1 (16-12-2014)

Licence

Identifiants

  • HAL Id : hal-01096027 , version 1

Citer

Jennifer Renoux, Abdel-Illah Mouaddib, Simon Le Gloannec. A Distributed Decision-Theoretic Model for Multiagent Active Information Gathering. Workshop on Multiagent Sequential Decision Making Under Uncertainty,, May 2014, Paris, France. ⟨hal-01096027⟩
298 Consultations
151 Téléchargements

Partager

  • More