Co-Fields: Towards a Unifying Approach to the Engineering of Swarm Intelligent Systems

摘要

Swarm intelligent systems, in which the paths to problem solving emerge as the result of interactions between simple autonomous components (agents or ants) and between them and their environment, appear very promising to develop robust and flexible software application. However, the variety of swarm-based approaches that have been proposed so far still lacks a common modeling and engineering methodology. In the attempt to overcome this problem, this paper presents a general coordination methodology in which swarm's components are simply driven by abstract computational force fields (Co-Fields), generated either by agents, or by the environment. By having agents be driven in their activities by such fields, globally coordinated behaviors can naturally emerge. Although this model still does not offer a complete engineering methodology, it can provide a unifying abstraction for swarm intelligent systems and it can also be exploited to formalize these systems in terms of dynamical systems whose behavior can be described via differential equations. Several example of swarm system modeled with Co-Fields are presented to support our thesis.