Robotic swarm cooperation by co-adaptation

  • Authors:
  • François-Michel De Rainville

  • Affiliations:
  • Université Laval, Québec, PQ, Canada

  • Venue:
  • Proceedings of the 14th annual conference companion on Genetic and evolutionary computation
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a framework for co-adapting mobile sensors in hostile environments to allow telepresence of a distant user. The presented technique relies on cooperative co-evolution for sensor placement. It is shown that cooperative co-evolution is able to find simultaneously the required number of sensors to observe a given environment and a configuration that is consistently better than other well know optimization algorithms. Moreover, it is presented that co-evolution is also able to quickly reach a new configuration when the environment changes.