Diagnostic model for identifying the technical condition of microprocessor systems with elements of artificial intelligence

The article proposes a diagnostic model for recognizing the technical condition of microprocessor systems, which is based on the ideas of artificial intelligence. An analysis of existing models and methods of recognizing the technical condition is carried out, the disadvantages and advantages are given. It has been proven that the most effective are discrete models, which are presented in the form of neural ensembles, and are nonlinear converters of analog information specified by a set of certain statistical and dynamic characteristics. At the same time, the formation of ensembles in the neural network corresponds to the memorization of images, signs, objects, etc. It is substantiated that the description of the technical state of microprocessor systems through the use of a discrete-event model, in which cause-and-effect viscous systems represent discrete random Markov chains in combination with an artificial neural network, opens up new opportunities for the intellectualization of the procedure for recognizing the technical state of microprocessor systems. In general, this article is a series of works related to the consideration of the main tasks of technical diagnostics, namely: assessment of the technical condition of microprocessor systems, as well as detection and localization of malfunctions, provided that the microprocessor systems are implemented on programmable logic integrated circuits and function in the conditions of internal and external adverse influences.