Models of Neuro-fuzzy Agents in Intelligent Environments

The set of agent-oriented information systems that interact in a changing structure and operation conditions, is considered as a distributed intellectual environment. Such a representation is corresponding in geographically distributed systems (region, city), corporate information systems of enterprises and organizations, social groups within the various territorial structures, distributed heterogeneous systems and networks territorial entities, enterprises and organizations. On the basis of known models of neuro-fuzzy agents are proposed a model of the fundamental intelligent agent (FIA), operating in a dynamic heterogeneous information environment and carrying out reception of input messages in a usual source language and fuzzy input language, the creation of messages in the inner language to perform reasoning and reflection which can be either conventional or fuzzy, and transmitting to the external environment accurate and fuzzy messages in appropriate languages. A classification of possible FIA models are constructed in distributed intelligent systems and environments. A behaviour model of the FIA, defined Post calculus axiom schemes, combining precise and fuzzy representation of variables and manipulate the internal states, base attributes and logical formulas in fuzzy logic and first-order predicate calculus are constructed.