A fitness-sharing based genetic algorithm for collaborative multi robot localization

  • Authors:
  • Francesco Bori;Andrea Gasparri;Stefano Panzieri

  • Affiliations:
  • Department of Computer Science, University of "Roma Tre", Via Vasca Navale, Roma, Italy;Department of Computer Science, University of "Roma Tre", Via Vasca Navale, Roma, Italy;Department of Computer Science, University of "Roma Tre", Via Vasca Navale, Roma, Italy

  • Venue:
  • IROS'09 Proceedings of the 2009 IEEE/RSJ international conference on Intelligent robots and systems
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, a novel genetic algorithm based on a "collaborative" fitness-sharing technique to deal with the Multi-Robot Localization problem is proposed. Indeed, the use of the fitness-sharing is twofold and competitive. It preserves the diversity among individuals during the space exploration process, thus maintaining evolutionary niches over time, and reinforces the best hypotheses by means of collaboration among robots, thus augmenting the selection pressure. Simulations by exploiting the robotics framework Player/Stage have been performed along with a proper statistical analysis for performance assessment.