Моделювання взаємодії нейронів живої нейромережі для задач технологій штучного інтелекту

Abstract

An author’s model of the process of controlling the propagation of an information signal in a living neural network is presented, which can be useful in the development of principles for controlling robot-technical complexes with artificial intelligence, as well as in the field of diagnosis and treatment of diseases of the central nervous system of a person.The new model for the interactionof neurons in a living neural network, considered in the article, explicitly takes into account the phenomenon of delay in the interaction time of a group of interconnected neurons, known in neurophysiology. To describe these processes, the mathematical apparatus of differential equations with delayed argument (DUZ) for the general case was first developed. The system of differentialequations describes the dependence of the dynamics of excitation in a group of neurons on the totality of their basic biochemical parameters: on a specific frequency of neuron oscillations; the amount of attenuation in time of these oscillations, the intensity of the action of one neuron on another. Thedependence of the intensity of the excitation conductivity in the neural network on the indicated biochemical characteristics can be taken from the results of experimental work.Ref. 6