A E Petrov¹ , A V Fedorov² , A V Kochegarov² , E N Lomaev³ and A P Preobrazhenskiy⁴

¹Department of APD, National Research Technological University "MISiS", 4, Lenina distr, Moscow, 119991, Russia

² Department of Scientific Investigations, Academy of the State Fire Service EMERCOM of Russia, 4, Borisa Galushkina st., Moscow, 129366, Russia

³ Leadership Department, LENProektInstallation, 1, Eniseyskaya st, Moscow, 129344, Russia

⁴ Information Systems And Technologies Department, Voronezh institute of high technologies, 73a, Lenina st, Voronezh, 394043, Russia

Abstract

Tensor network models represent the processes and structure of complex systems, including dangerous technological objects of oil refineries. Complex systems tensor method research based on duality invariance. It is the sum constancy of the metric tensors dual networks whenever there is a change in structure. Dual subspaces closed and unclosed paths are responses to external and internal influences, which do not depend on each other. As an example of a network model is considered the atmospheric column oil-refining plant, which is used to organize the work of training stand. Thermal energy flows are presented in a mesh network, and massive flows of mechanical energy - in a junction network. The amount of mechanical and thermal energy flows allows you to determine the load on the nodes of the unit. The results are applied to predict dangerous situations on refining facilities to assess changes in restructuring processes and the risk of accidents. Calculations, such as heat and mass flow change if you change the control actions, disable installation elements and assess change critical settings.