Naslov
Prediction of natural gas consumption by neural networks

Autori
Šebalj, Dario ; Mesarić, Josip ; Pap, Ana

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the 76th International Scientific Conference on Economic and Social Development Development – "Building Resilient Society" / Mišević, P. ; Kontić, Lj. ; Galović, T. - Zagreb : VADEA, 2021, 248-258

Skup
76th International Scientific Conference on Economic and Social Development: "Building Resilient Society"

Mjesto i datum
Zagreb, Hrvatska, 17.12.2021. - 18.12.2021

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Recenziran

Ključne riječi
Algorithms ; Energy ; Natural gas ; Multilayer Perceptron ; Machine learning

Sažetak
Due to its environmental benefits, natural gas has become one of the most popular energy sources. Natural gas is the third largest energy source in 2020, after oil and coal, accounting for nearly 25%. The consumption of natural gas has been increasing in recent years, except for last year when consumption decreased by 2.3%. The aim of this paper is to present a neural network model (using Multilayer Perceptron algorithm) that could predict natural gas consumption on an hourly basis. The dataset consists of hourly natural gas consumption data obtained from natural gas supplier and distributor, and meteorological data. There have been many studies in which researchers have attempted to predict gas consumption, and the accuracy of these models is important for decision making, especially for gas nominations (gas orders). The results show that the statistical correlation between the actual and predicted values is very high, but the relative absolute error and root relative squared error are about 25% which cannot be considered satisfactory for this type of prediction. The comparison between the actual and predicted values shows that the model appears to be good at predicting gas consumption in the winter months, but predicts lower values than actual ones for the summer months.

Izvorni jezik
Engleski

Znanstvena područja
Ekonomija, Informacijske i komunikacijske znanosti

POVEZANOST RADA