Functional structure of the cohesive hybrid intelligent multiagent system

The effectiveness of the hybrid and synergetic artificial intelligence systems is achieved, mainly, by the proper organization of the interaction of their elements and not by increasing the complexity or intelligence of the latter. In the case of hybrid intelligent multiagent systems, simulating teams of experts solving problems “at round-table,” for the effective organization of interaction between agents, the conditions for the consistency of their goals and domain models, the unity of the protocol for solving the problem, and the compatibility of message transfer languages should be fulfilled. These conditions are especially relevant when a hybrid intelligent multiagent system is built from agents developed by various independent teams. In this case, agents should be able to independently coordinate goals, domain models and develop a protocol to solve the problem posed to them within the framework of a new class of intelligent systems, namely, a cohesive hybrid intelligent multiagent system. This article discusses the functional structure of such a system.