A New Dynamic Model for a Multi-Agent Formation

  • Authors:
  • Daniel Gonçalves;Pedro Machado;Ana Paiva

  • Affiliations:
  • -;-;-

  • Venue:
  • IBERAMIA '98 Proceedings of the 6th Ibero-American Conference on AI: Progress in Artificial Intelligence
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

Many systems have been already developed concerning agent teams in a world with obstacles. One of the problems of such systems lies on how to maintain a pre-defined formation when we have several agents moving in the world.In this paper we defend that, in order to have a robust and realistic system, a control model that includes the notions of mass and acceleration must be used. To prove that, we developed a control system based on the classic mechanical physics, which is a force-based model. From the results obtained we can see that, although some problems arise when using such realistic kind of model, they are solvable and the quality of the simulations performed by the system is significantly better than the simulations obtained using other control models.