Multiagent meta-model for strategic decision support

摘要

This research describes a meta-model which can generate multiagents system (MAS) of strategic decision support. These MAS profit from new concepts of bargaining with learning to determine the cooperation between agents. They answer an insufficiency of the concept of bargaining in the game theory which is proved in this article. The original framework of the meta-model is characterized by three essential functions provided with a technique of learning: the search for typical plan of decisions adapted to the problem and the search for the coalitions in the spatial and temporal dimension. It allows an adaptation to all the types of strategic decision support.