THE USE OF ARTIFICIAL INTELLIGENCE IN THE ENERGY LOSS METERING SYSTEM  

Dmitry N. Frantasov, Candidate of technical sciences, associate professor of the sub-department of applied informatics, Samara State University of Economics (141 Sovetskaya Armiya street, Samara, Russia) frantasov@mail.ru
Yulia V. Kudryashova, Senior lecturer of the sub-department of higher mathematics, Volga State University of Railway Transport (2V Svobody street, Samara, Russia) yu.kudryashova@samgups.ru
Еlena V. Voronina, Lecturer of the sub-department of applied informatics, Samara State University of Economics (141 Sovetskaya Armiya street, Samara, Russia) voroninae.v@sseu.ru

Background. The aim of the research is to reduce electricity losses in the power system of railway transport. Losses are defined as the imbalance between the supplied and consumed electricity recorded by means of commercial electricity metering. At the moment, there are no technical and software tools for analyzing the losses structure and causes of their occurrence. Materials and methods. To achieve this goal the main factors that lead to a significant increase in electricity losses have been identified. The factors for all automated systems for commercial metering of electricity used in railway transport have been determined. Results. A block diagram of an information and measurement system for metering electricity losses has been developed, which allows using the equipment in operation without replacing or modernizing, which contributes to getting new technical capabilities. A method is proposed for the intellectualization of the process of classification of factors causing the growth of unnormalized energy losses, based on artificial neural networks. Conclusions. The intelligent unit allows you to replace the person who makes organizational and technical decisions to minimize the consequences of emergency situations leading to an increase in non-standardized losses. This solution is applicable in departments without qualified specialists. The results of training an artificial neural network are considered, the main parameters of the efficiency of the information and measurement system for recording losses at a real object of railway transport are determined.

power losses, power imbalance, smart power systems, artificial neural networks

Frantasov D.N., Kudryashova Yu.V., Voronina E.V. The use of artificial intelligence in the energy loss metering system. Modeli, sistemy, seti v ekonomike, tekhnike, prirode i obshchestve = Models, systems, networks in economics, technology, nature and society. 2024;(3):116–125. (In Russ.). doi: 10.21685/2227-8486-2024-3-10