The article describes general approaches to creating an intelligent system for monitoring and diagnosing the operability of energy supply facilities. The general concept of the adaptive-predictive analysis system and the construction of an artificial neural network for its use in the predictive module for predicting the type and time of failure occurrence is given. The basic principles of training a neural network for recognizing various types of failures are also given. Critical remarks of the concept of creating a digital twin of such a complex object for modeling as energy-generating equipment are given.

This content is only available via PDF.
You do not currently have access to this content.