Primena mašinskog učenja u predviđanju energetskih performansa toplifikacionog sistema
##plugins.themes.bootstrap3.article.sidebar##
##plugins.themes.bootstrap3.article.main##
http://orcid.org/0009-0001-9160-6379
Ivan Ćirić
http://orcid.org/0000-0003-0430-8937
Marko Ignjatović
http://orcid.org/0000-0001-7205-7987
Apstrakt
U ovom radu istraživane su mogućnosti predviđanja energetskih performansi toplifikacionog sistema Mašinskog fakulteta u Nišu primenom nadgledanog mašinskog učenja. Kontrola rada top- lifikacionog sistema Mašinskog fakulteta u Nišu se odvija automatski. Praćenje energetskih perfor- mansi obavlja se pomoću SCADA sistema ali odluke o radu, koje se tiču unapređenja sistema u smislu uštede energije a samim tim i smanjenja troškova, donosi operater toplane. Predložena predviđanja u radu zasnivaju se na primeni veštačkih neuronskih mreža nad skupom energetskih indikatora preuzetih iz SCADA sistema toplane Mašinskog fakulteta u Nišu. Predstavljeni su rezultati predikcije utrošene toplotne energije u posmatranom vremenskom intervalu koji obuhvata period od 15 dana, dobijeni primenom različitih alogiritama neuronskih mreža u softverskom alatu Matlab a u cilju da se pokaže da su korišćeni algoritmi neuronskih mreža sposobni da, sa dovoljnom preciznošću, razumeju kompleksan sistem daljinskog grejanja kako bi se moglo omogućiti korišćenje nekih od prikazanih modela predviđanja za detekciju anomalija u radu toplifikacionog sistema Mašinskog fakulteta u Nišu.
##plugins.themes.bootstrap3.article.details##
Reference
[2] Milan Zdravković, Ivan Ćirić, Marko Ignjatović, Towards explainable AI-assisted operations
in district heating systems, IFAC PapersOnLine 54-1 (2021) 390–395, Niš, 2021.
[3] Miloš B. Simonović, Vlastimir D. Nikolić, Emina P. Petrović and Ivan T. Ćirić, Heat load prediction of small district heating system using artificial neural network, Thermal Science, Vol.
20 (2016), Suppl. 5, pp. S1355-S1365
[4] Ivan Ćirić, Marko Ignjatović, Mirko Stojiljković, Dušan Stojiljković, Milan Gocić, Milica
Ćirić, Intelligent Heat Demand prediction for Advanced District Heat Plant Control, 10th Inter-
national Conference on Information Society and Technology, Belgrade, Serbia, 2020.
[5] Simonović, M. B., Artificial Neural Network Application for Short-Term Prediction and Anal-
ysis of District Heating Systems, Ph.D. thesis, University of Niš, Niš, Serbia, 2016.
[6] Davide Quaggiotto, Jacopo Vivian, Angelo Zarrella, Management of a district heating net- work using model predictive control with and without thermal storage, Optimization and Engi-
neering, Vol. 22 (2021), Issue 4, pp. 1901–1931.
[7] Madhuranthakam, C., Nigam, A., Pathak, P., A review of PID control, tuning methods and
applications, International Journal of Dynamics and Control, Vol. 9 (2021), Issue 1, pp. 88–
105.
[8] Bukhari, A., S. H. Yusoff, M. S. F. M. Yunus, N. S. I. Razali, Virtual Power Plant Management
Using PID Controller, Journal of Renewable Energy and Power Systems, Vol. 9 (2021), Issue,
pp. 818–827.
[9] Maryniak, A., M. Banaś, P. Michalak, J. Szymiczek, Forecasting of Daily Heat Production in
a District Heating Plant Using a Neural Network, Energies, Vol. 17 (2024), issue 17, pp. 4369.
[10] Hae-Won Uh, Lucia Klarić, Ivo Ugrina, Gordan Lauc, Age K. Smilde and Jeanine J. Houwing-Duistermaat, Choosing proper normalization is essential for discovery of sparse gly-
can biomarker, Molecular Omics, Vol. 16 (2020)
[11] B. Chramcov, Heat demand forecasting for concrete district heating system, Int. J. Math. Model.
Methods Appl. Sci., Vol. 4, no. 4, pp. 231– 239, 2010.
[12] Petrović, M., Basics of Artificial Neural Networks and the Importance of Their Application,
Proceedings of the Faculty of Civil Engineering, Vol. 20 (2011), pp. 47–55.
[13] Đozić, D., The Use of Artificial Neural Networks for Predicting Behavior and Managing Com-
plex Energy Systems, Ph. D. thesis, University of Novi Sad, Novi Sad, Serbia, 2020.
[14] J. Xie, H. Li, Z. Ma, Q. Sun, F. Wallin, Z. Si, J. Guo, Analysis of Key Factors in Heat Demand Prediction with Neural Networks, Proceedings of 8th International Conference on Applied En-
ergy, ICAE2016, 8-11 October 2016, Beijing, China, pp. 2965-2970