Method for controlling the cognitive systems functioning in non-equilibrium stability mode

Abstract

Cognitive systems as self-organizing systems have a number of special properties, including the ability to function in a mode of non-equilibrium stability, which, in turn, provides the possibility of self-organization of the system by transforming its structure. In this case, one of the most important is the task of monitoring the functioning of the system in the mode of non-equilibrium stability. The development of a method for controlling the functioning of cognitive systems in the mode of non-equilibrium stability is based on the previously proposed model of non-equilibrium stability, which assumes a probabilistic assessment of the effectiveness of the structural elements of the system. To assess the effectiveness of structural elements, a method based on the principle of comparative preference is proposed. The method is implemented using an appropriate algorithm, which made it possible to numerically simulate it in order to determine the accuracy of evaluating the effectiveness of structural elements. Modeling has shown that the accuracy of the assessment depends on the number of structural elements and on the number of performance indicators. As a result of the simulation, the areas of non-equilibrium stability of the system functioning were determined for a different number of structural elements and the probability of their effectiveness. The proposed control method makes it possible to determine the position of the area of non-equilibrium stability in the phase space of the system functioning and, on the basis of this, evaluate both the current state of the system and predict the possible state of the system when it leaves the region of non-equilibrium stability.