Making predictions for overhead power lines applying predictive analytics

Introduction. Currently, failures are one of the main and urgent problems for the normal functioning of power supply systems, they are at first glance unpredictable, however, in addition to accidental, there are failures that occur according to some regularity. Power transformers, electric machines and transmission lines of different voltage levels are considered the main electrical equipment. It should be noted that the general trend is a high percentage of wear and tear, moral and physical aging of the main and auxiliary electrical equipment.

Goal - the application of methods of mathematical analysis and predictive analytics to predict the number of failures of electrical equipment and the development of an algorithm for calculating forecasting to assess damages in the future.

Materials and methods. The applied methods are based on approximation and on extrapolation of the data of electrical equipment failures for 2011-2021 in the conditions of the power supply system of Magnitogorsk energy hub. For convenience of work with mathematical apparatus were used software programs “Matlab” in the format of directed vectors and utilities “Curve fitting”, as well as software “Microsoft Excel”.

Results and discussion. The article proposes a methodology for predicting failures for the future period, which can allow, given a sufficient amount of input data, to obtain the estimated number of failures for the object under study.

Conclusion. According to the results of the conducted research the methodology for short-term forecasting of the number of failures was developed, it was also applied to real statistical data and showed acceptable reliability and adequacy. In the future, similar forecasts can be performed for various types of electrical equipment in the conditions of urban and industrial networks for future planning of budgetary funds allocated for additional diagnostics, replacement and preventive maintenance as of the current moment, which will prevent the scale of accidents, reduce their number, which will lead to increased reliability of the power supply system of industrial and urban consumers.

1. Gockenbach E, Zhang S, Liu C et al. Predicting the service life of power transformers by condition monitoring. Elektroener- giya. Peredacha i raspredelenie = Electricity. Transmission and Distribution. 2014;4(25):82-88. (In Russ.).

2. Ilyichev VYu, Yurik EA. Application of the predictive analytics methods for signal processing from the rotor machine speed sensors. Nauchnoe obozrenie. Tekhnicheskie nauki = Scientific review. Technical Sciences. 2019;1:22-26. (In Russ.).

3. Odintsov KE, Rotanova YN, Karandaeva OI et al. Methodology for predicting the residual life of electrical equipment in operation / Izvestiya Tul'skogo gosudarstvennogo universiteta = Izvestia Tula State University. Technical sciences. 2010;3- 1:192-198. (In Russ.).

4. Kondrashova YuN, Tretyakov AM, Shalimov AV. Evaluation of the impact of the development of regulations for maintenance and repair of production equipment in order to minimize costs and losses of ferrous metallurgy enterprise. Vestnik Magnitogorskogo gosudarstvennogo tekhnicheskogo universiteta im. GI Nosova = Bulletin of Magnitogorsk State Technical University named after GI Nosov. 2022;(20(1)):112-120. https://doi.org/10.18503/1995-2732-2022-20-1-112-120. (In Russ.).

5. Methods of interpolation and approximation. Portal.tpu. Available from: https://portal.tpu.ru/SHARED/m/MBB/uchebnaya_rabota/Model/Tab/Interp_app.pdf [Accessed 14 June 2023]. (In Russ.).

6. Strashko IV, Kravchenko ТС. Essence of extrapolation methods. Studopedia. Available from: https://studopedia.ru/17_38085_sushchnost-metodov-ekstrapolyatsii.html [Accessed 14 June 2023]. (In Russ.).