Modeling of Reasoning a Cognitive Agent with Significant Time Restrictions

Abstract

An integrated logical system that combines the concepts of active logic and logical programming has been built.. An approach to constructing a paraconsistent declarative semantics based on the concept of active logic has been introduced. The method of representing agent’s temporal non-monotonous reasoning by active temporal logic has been proposed. The temporal granulation technique in logical system to formalize meta-reasoning has been suggested. Taking into account that the agent has to make decisions under the lack of time. the problem of the decision quality arises. In this context, it is useful to take branching time logics that allows us to infer various consequences of agent’s decision. A subclass of such logics oriented to real-time systems applications has been considered. In general, the proposed methods and algorithms provide the conceptual and algorithmic bases for developing new generation intelligent systems able to function in the case of hard real-time.