Розробка моделі цифрового двійника для системи моніторингу технічного стану енергоблоку атомної електростанції

This study investigates the process of continuous monitoring and control of the technological process parameters at a nuclear power plant power unit using a digital twin model implemented on the basis of a system-cluster approach. The task addressed relates to the lack of a comprehensive, integrated system for continuous monitoring of the technical condition of the power unit, capable of collecting, processing, and analyzing information from sensors and diagnostic subsystems in real time. It is proposed to model the state of the technological equipment at a power unit using a digital twin and a system-cluster approach. Within the framework of the study, a system-cluster architecture of a digital twin was designed, which reflects the interaction among the physical, analytical, and control levels of the power unit at a nuclear power plant. The work involved processing the technological process parameters coming from a sensor network with a frequency of 1–2 Hz, with a processing delay of no more than 1–3 s. The proposed information-fractal criterion provided an increase in the sensitivity of pre-accident detection by 15–25% compared to conventional methods, as well as made it possible to identify complex operating modes in the range of . The results have made it possible to solve the set tasks by integrating multi-scale fractal analysis, cluster organization of technical systems, and self-similarity modeling. The practical implementation of the digital twin has proven its capability to detect changes in the structure of the technological process with diagnostic accuracy at the level of 85–92%. The implementation of the digital twin model in the information-control systems of the power unit makes it possible to increase the reliability, safety, and efficiency of control.