Tasks-based MAS model on the granular computing framework

Granularity theory is a new soft computing tool, which can deal with vagueness and uncertainty efficiently. In this paper, we introduce the MAS model based on the cooperative tasks on the granular analysis and computing framework. We discuss the necessity of agent in MAS and the minimum reduction based on the given tasks, and then put forward the criterion to classify them. Also, we propose the notion of significant and cooperative degree of MAS, and conclude the matching algorithm with respect to the cooperative tasks.