Emergence of Behaviours in Natural Phenomena Agent-Simulation

Pierre Marcenac

Abstract:

This article presents some results of the GEAMAS framework, a generic architecture to develop multi-agent simulations of complex systems. The architecture of GEAMAS is based on the definition of three levels, each successive level representing a higher level of abstraction. The architecture has been studied to distribute the intrinsic complexity of the system in three levels. With that aim, an agent model has been defined, through the representation of interactions, behaviours and dynamic evolution capabilities. Evolution in multi-agent systems can be classified in several categories and could address agents knowledge, agents structure (self-organisation) and auto-adaptation.

The main issue tackled in this paper is auto-adaptation - that is generation of behaviours from the study of local interactions between agents and the environment. Some key issues associated with the understanding and representation of such emerging behaviours in multi-agent systems are introduced.

Keywords:
Emergence, Local Interactions, Complex Systems Modelling, Agent-Oriented Knowledge Representation, Simulation Applications.

• Introduction
• Distributing Complexity in Reactive and Cognitive Agents
  • A Society of Agents
  • Cognitive and Reactive Agents in the GEAMAS Architecture
  • Reactive Agents
• Emergence of Complex Behaviours from Local Interactions
  • Emergence in GEAMAS
  • Implementation of Emergence in GEAMAS
• Discussion and Concluding Remarks
• Acknowledgements
• References
• About this document ...

Complexity International (1996) 3