Neural network monitoring and predictive control of the reliability and safety of gas distribution networks using deep learning algorithms

Abstract

The work deals with the development of an algorithmic, mathematical complex based on neural networks of deep learning to assess and predict the reliability and safety of gas distribution equipment based on diagnostic data. The article describes the features of creating a predictive control system within the framework of the CI-TREMS complex. The work describes the algorithms and presents an example of the implementation of the system on one of the nodes of the gas distribution network.