A Generic Model for Swarm Intelligence and Its Validations

Abstract

The modeling of emergent swarm intelligence constitutes a major challenge and it has been tacked in a number of different ways. However, existing approaches fail to capture the nature of swarm intelligence and they are either too abstract for practical application or not generic enough to describe the various types of emergence phenomena. In this paper, a contradiction-centric model for swarm intelligence is proposed, in which individuals determine their behaviors based on their internal contradictions whilst they associate and in-teract to update their contradictions. The model hypothesizes that 1) the emergence of swarm intelligence is rooted in the development of individuals’ internal contradictions and the interactions taking place between in-dividuals and the environment, and 2) swarm intelligence is essentially a combinative reflection of the configu-rations of individuals’ internal contradictions and the distributions of these contradictions across individuals. The model is formally described and five swarm intelligence systems are studied to illustrate its broad applica-bility. The studies confirm the generic character of the model and its effectiveness for describing the emergence of various kinds of swarm intelligence; and they also demonstrate that the model is straightforward to apply, without the need for complicated computations.