Development of Verification Model for Intelligence of Multiagent Systems.

The development of distributed information systems essentially relies on the complex application of Web services technologies, cloud, “green”, GRID, multi-agent and many others. In many of these technologies, intellectualization is increasingly manifested, which is clearly inherent, for example, in semantic and intelligent Web services, multi-agent systems. Intelligence, often defined as a set of properties of knowledge and mechanisms for their formation, requires its own ways of formalization and implementation, which is also reflected in the problems of analysis and synthesis of intelligent systems, including the problems of their identification and verification. This paper consider the fuzzy and component development of a model for verifying the intelligence of multi-agent systems based on the previously proposed model for verifying the intelligence of multi-agent systems, which, in turn, is based on behavioral experiments for extended Petri nets. The features of the proposed model for verifying intelligence are the presentation of fuzzy production models of all components of intelligence for parallel, associative-evolutionary search for known identifying primitives and recognition of fragments of system behavior based on them, the construction of new identifying primitives as part of a three-level model of new knowledge, as well as a fuzzy inference mechanism, as the basis for obtaining new knowledge, and for identifying and verifying the behavioral properties of the system. The intelligence verification model makes it possible to reduce the computational complexity of the behavioral verification of multi-agent systems and increase its flexibility.