The Use of Machine Learning Methods to Transform the Parameters of the Gas Turbine Operation for the Diagnosis of a Turbocharger

Since 2017, domestic monitoring and predictive diagnostics systems for power equipment have been developing. The Fourth Industrial Revolution contributed to the development of these systems “Industry 4.0”, as well as widespread digitalization and an increase in a variety of control and measuring devices, which means an increase in statistical data sets, without which the development of these systems would be impossible. In addition, with the universal course of import substitution, Russian gas turbines have actively begun to be developed and implemented, requiring increased attention from diagnostic systems for normal operation and operation control. This paper describes an algorithm for transforming the parameters of the gas turbine operation, built on the basis of machine learning. This conversion is necessary for further diagnostics of the gas turbine turbocharger and identification of its main malfunctions.