Тип публикации: статья из журнала
Год издания: 2023
Идентификатор DOI: 10.1088/1755-1315/1231/1/012074
Аннотация: <jats:title>Abstract</jats:title> <jats:p>Information technologies have entered all spheres of human life. Now they are an integral part of the modern world. These changes also affected the anti-crisis management system. The development of end-to-end technologies and the use of artificial intelligence allows for a better assessmentПоказать полностьюand forecasting of risks. The methods used for forecasting the inflow to the site of a hydroelectric power station were analyzed in this paper. The data necessary to obtain a forecast were reviewed. An approach related to forecasting the inflow to the site of a hydroelectric power station using neural network forecasting methods is demonstrated. Using the python programming language, a neural network model was developed to predict the inflow to the hydroelectric station site. A recurrent network with a long short-term memory LSTM using 64 neurons was chosen as the network architecture. The data array for forecasting was obtained from the data of operational monitoring of the flood situation and archival data on flooding of the territory over the past 13 years, taken from the database of the Yenisei Basin Water Administration of the Federal Agency for Water Resources and the Central Siberian UGMS. The input parameters were selected empirically. When comparing real data on flow with the predicted ones, the error was 17%.</jats:p>
Журнал: IOP Conference Series: Earth and Environmental Science
Выпуск журнала: Т. 1231, № 1
Номера страниц: 012074
ISSN журнала: 17551307
Издатель: IOP Publishing