FORECASTING THE LOAD ON THE POWER GRID AS A WAY TO EFFECTIVELY MANAGE THE CONSUMPTION OF ELECTRICAL ENERGY
Keywords:
energy saving, forecasting, modeling, regression, correlation, machine learning, load on the power gridAbstract
Energy management systems (EMS System) are rapidly developing at this moment. They allow reducing the financial costs of enterprises for electric energy. This cost reduction is achieved due to the optimal management of the company's power supply system. This paper presents the results of intelligent analysis of data on electricity consumption on the example of an office building located in the city of Moscow. This building has four inputs for electrical energy, each of which has a digital consumption meter that records readings every 30 minutes in an automated commercial electricity metering system. In addition, an electrical energy storage system is installed in the object under consideration, for the formation of optimal control actions on which it is required to perform qualitative forecasting of the electrical load. This will allow industrial electric energy storage to be used at the right time periods to reduce the electric energy consumed from the grid during peak hours (when using a two-part tariff) or in the peak zone of the day (when using a differentiated tariff for electric energy). In the course of the study, the initial data was preprocessed and supplemented with signs from external sources: weather data, production calendar. To assess the importance of features, a correlation matrix was constructed, as a result of the analysis of which the need to remove some factors from the model as insignificant for the target variable was revealed. In addition, the analysis of the correlation matrix confirmed the absence of intercorrelation. To assess the quality of the models, a set of basic statistical indicators of forecast accuracy was selected, on the basis of which a comparative analysis of the forecast quality of the developed models was carried out. The result of the study was the choice of a machine learning model based on the regression model of extreme gradient boosting (XGBRegressor) and allowing to obtain the most reliable forecast of the load on the building's power grid. The study was carried out according to the data of the load of electricity on an office building, however, the results of the study can be adapted for other objects.
References
Gale Boyd E. Mark Curtis, Su Zhang. Impact of Strategic Energy Management Practices on Energy Efficiency: Evidence from Plant-Level Data. Summer Study on Energy Efficiency in Industry 2021. P. 0023123.
Паламарчук А.Г. Анализ современного состояния энергосбережения в российской промышленности // Научные труды Вольного экономического общества России, 2020. С. 270–282. / Palamarchuk A.G. Analysis of the current state of energy saving in the Russian industry. Scientific works of the Free Economic Society of Russia, 2020. P. 270-282.
Попадько Н.В., Найденова В.М. Энергосбережение и повышение эффективности как вектор развития мирового энергетического комплекса // Инновации и инвестиции № 5, 2020. С. 91–95. / Popadko N.V., Naidenova V.M. Energy conservation and efficiency improvement as a vector of development of the world energy complex. Innovations and Investments. № 5, 2020. P. 91–95.
Исламова В.М., Мустафин Т.Р., Кантемиров И.Ф. Комплексный подход к использованию вторичных энергоресурсов на компрессорной станции // Транспорт и хранение нефтепродуктов и углеводородного сырья, 2020. С. 27–31. / Islamov, V.M., Mustafin T.R., Kantemirov I.F. A comprehensive approach to the use of secondary energy resources at compressor stations. Transport and storage of petroleum products and hydrocarbons, 2020. P. 27–31.
Жаворонкова Н.Г., Шпаковский Ю.Г. Энергетическая стратегия 2035: правовые проблемы инновационного развития и экологической безопасности // Вестник Университета имени О.Е. Кутафина, 2020. С. 31–47. / Zhavoronkova N.G., Shpakovskiy Yu.G. Energy strategy 2035: legal problems of innovative development and environmental security. Bulletin of the O.E. Kutafin University, 2020. P. 31–47.
Фаустова И.Л. Опыт эффективного управления энергосбережением в промышленности развитых стран // Экономический анализ: теория и практика, 2010. С. 59–64. / Faustova I.L. Experience of effective energy conservation management in the industry of developed countries. Economic Analysis: Theory and Practice, 2010. P. 59–64.
Соколов А.А. Методы поддержки принятия решений в системах обеспечения энергетическими ресурсами на машиностроительных предприятиях. Автореферат диссертации, Волгоград. 2019. 20 с. / Sokolov A.A. Methods of decision support in energy resources supply systems at machine-building enterprises. Abstract of dissertation, Volgograd. 2019. 20 p.
Мартынюк М.В. Модели и алгоритмы интеллектуального управления параметрами регулирующих устройств в цифровых электросетях. Автореферат диссертации, Нижний Новгород, 2019. 23 с. / Martynyuk M.V. Models and algorithms of intelligent control of parameters of regulating devices in digital electrical networks. Abstract of dissertation, Nizhny Novgorod, 2019. 23 p.
Зарипова В.М., Квасова В.О., Петрова И.Ю. Децентрализованный рынок электроэнергии – новые требования к цифровизации бизнес-процессов // Инженерно-строительный вестник Прикаспия, 2020. С. 98–104. / Zaripova V.M., Kvasova V.O., Petrova I.Yu. Decentralized electricity market – new requirements for digitalization of business processes. Engineering and Construction Bulletin of the Caspian Sea, 2020. P. 98–104.
Клюев Р.В., Гаврина О.А., Хетагуров В.Н., Зассеев С.Г., Умиров Б.З. Прогнозирование удельного потребления электроэнергии обогатительной фабрики // Горный информационно-аналитический бюллетень (научно-технический журнал), 2020. №11-1. С. 135–145. / Klyuev V.R., Gavrila O.A., Khetagurov V.N., Zaseev S.G., B.Z. Umirov Prediction of specific consumption of electric energy concentrator. Mining informational and analytical Bulletin (scientific and technical journal), 2020. № 11-1. P. 135–145.
Adel Naji, Badr Al Tarhuni, Jun-Ki Choi, Salahaldin Alshatshati, Seraj Ajena. Toward cost-effective residential energy reduction and community impacts: A data-based machine learning approach // Energy and AI. 2021. Vol. 4. P. 37–55.
Shintaro Ikeda, TatsuoNagai. A novel optimization method combining metaheuristics and machine learning for daily optimal operations in building energy and storage systems // Applied Energy. 2021. Vol. 289. P. 67–82.
Cheng Zhoua, Xiyang Chen. Predicting China’s energy consumption: Combining machine learning with three-layer decomposition approach // Energy Reports. 2021. Vol. 7. P. 5086–5099.
Wenqiang Jing, Junqi Yu, Wei Luo, Chujun Li, XinYi Liu. Energy-saving diagnosis model of central air-conditioning refrigeration system in large shopping mall // Energy Reports. 2021. Vol. 7. P. 4035–4046.
Dominik Flick, Claudio Keck, Christoph Herrmann, Sebastian Thiede. Machine learning based analysis of factory energy load curves with focus on transition times for anomaly detection // Procedia CIRP. 2020. Vol. 93. P. 461–466.
