Registration Number
0224U033293,
(0123U101852)
, R & D reports
Title
Intelligent information technology for proactive management of energy infrastructure in conditions of risk and uncertainty
popup.stage_title
Розробка методів прогнозування для інформаційної технології проактивного управління енергетичною інфраструктурою
Head
Shendryk Vira V., Кандидат технічних наук
Registration Date
24-12-2024
Organization
Sumy State University
popup.description1
The project is devoted to the development of intelligent information technology for proactive management of energy infrastructure in conditions of risk and uncertainty, which will ensure the flexibility of functioning and sustainable development of the energy industry, as one of the most critical complex technical systems. It is planned to obtain new applied scientific and practical results in the form of methods and tools of intelligent data analysis and artificial intelligence to ensure operative decision-making in energy management, considering their current state and potential in accordance with configuration, functionality, and availability of resources. The aim of the project is to obtain new applied knowledge and practical development of intelligent distributed decision support systems for proactive management of energy infrastructure based on artificial intelligence and distributed information processing methods. In addition, it is planned to expand knowledge regarding the methods of determining the needs of system users, methods of data visualization in a user-friendly form, methods of forecasting peak loads, and planning the balancing of energy networks in accordance with the current configuration, functional capabilities, and availability of resources. This is complex research aimed at developing a set of new methods of building new information technologies and modern methods of intellectual data analysis using intelligent systems built on the knowledge and technologies of artificial intelligence and edge computing
popup.description2
A study of time series forecasting methods for the subject area of renewable energy, as well as the influence of the duration of the forecasting interval and input data on the accuracy of forecasting models, was conducted. As a result, data structures were determined and data sets were formed for implementing tasks of forecasting the generation and consumption of electricity from renewable sources. Experiments to study the effectiveness of various forecasting methods were conducted. Machine learning models using neural networks were developed for hourly forecasting of electricity consumption by different types of consumers, including forecasting peak loads on the electrical network. Models of electricity generation from solar panels were also developed. A study was conducted on the adaptation of the developed forecasting models for use on mobile devices, which will allow for their widespread use in the future, reducing the cost of performing calculations. Software was developed, which was used to test models of consumption and generation of electricity from renewable sources. An integrated information environment has been created to support forecasting tasks, which includes a database of historical consumption indicators, modules for data preprocessing and normalization, a file repository of forecasting models, as well as user interfaces
Product Description
popup.authors
Bohachov Maksym V.
Boiko Olha V.
Bratushka Larysa M.
Grabina Kateryna V.
Zakharchenko Ivan P.
Kinshakov Eduard V.
Komin Anton S.
Nahornyi Volodymyr V.
Oskin Bohdan V.
Pavlenko Petro М.
Parfenenko Yuliia V.
Titariev Artem М.
Tymchuk Serhii O.
Kholiavka Yevhen Р.
Chychykalo Yevhenii А.
Shendryk Vira V.
Shepelev Dmytro O.
popup.nrat_date
2024-12-24
