On A General Framework To Study Cooperative Systems

A collection of many systems that cooperatively solve an optimization problem is considered. The consideration aims to determine criteria such that the systems as a whole show their best performance for the problem. A general framework based on a concept of structural complexity is proposed to determine the criteria. The main merit of this framework is that it allows to set up computational experiments revealing the criteria. In particular, the experiments give evidence to suggest that criteria of best performance are realized when the structural complexity of cooperative systems equals the structural complexity of the optimization problem. The results of the paper could give a new perspective in the developing of optimization methods based on cooperative systems.