Physics informed neural networks for 1D flood routing
Autori
Bojović, FilipMilašinović, Miloš
Jovanović, Branka
Krstić, Lazar
Stojanović, Boban
Ivanović, Miloš
Prodanović, Dušan
Milivojević, Nikola
Konferencijski prilog (Objavljena verzija)
Metapodaci
Prikaz svih podataka o dokumentuApstrakt
Machine learning methods have been widely and successfully applied in hydrological problems. Most of the methods, such as artificial neural networks, have been focused on estimating hydrological data based on observation over time. Even though these models provide good results, it can be observed that results become unreliable when the training dataset is small or when input data is significantly out of range compared to the training data. Therefore, a new approach is presented, in which artificial neural networks are trained to satisfy physical laws. This is conducted by a novel method called physics-informed neural networks (PINNs), in which physical principles are embedded in a custom loss function. This paper presents the application of physics informed neural networks for solving 1D flood wave propagation in open channels. The research has shown promising results.
Ključne reči:
physics informed neural networks / flood wave propagation / loss functionIzvor:
1st Serbian International Conference on Applied Artificial Intelligence (SICAAI), 2022Finansiranje / projekti:
- Ministarstvo nauke, tehnološkog razvoja i inovacija Republike Srbije, institucionalno finansiranje - 200122 (Univerzitet u Kragujevcu, Prirodno-matematički fakultet) (RS-MESTD-inst-2020-200122)
Kolekcije
Institucija/grupa
GraFarTY - CONF AU - Bojović, Filip AU - Milašinović, Miloš AU - Jovanović, Branka AU - Krstić, Lazar AU - Stojanović, Boban AU - Ivanović, Miloš AU - Prodanović, Dušan AU - Milivojević, Nikola PY - 2022 UR - https://grafar.grf.bg.ac.rs/handle/123456789/2771 AB - Machine learning methods have been widely and successfully applied in hydrological problems. Most of the methods, such as artificial neural networks, have been focused on estimating hydrological data based on observation over time. Even though these models provide good results, it can be observed that results become unreliable when the training dataset is small or when input data is significantly out of range compared to the training data. Therefore, a new approach is presented, in which artificial neural networks are trained to satisfy physical laws. This is conducted by a novel method called physics-informed neural networks (PINNs), in which physical principles are embedded in a custom loss function. This paper presents the application of physics informed neural networks for solving 1D flood wave propagation in open channels. The research has shown promising results. C3 - 1st Serbian International Conference on Applied Artificial Intelligence (SICAAI) T1 - Physics informed neural networks for 1D flood routing UR - https://hdl.handle.net/21.15107/rcub_grafar_2771 ER -
@conference{ author = "Bojović, Filip and Milašinović, Miloš and Jovanović, Branka and Krstić, Lazar and Stojanović, Boban and Ivanović, Miloš and Prodanović, Dušan and Milivojević, Nikola", year = "2022", abstract = "Machine learning methods have been widely and successfully applied in hydrological problems. Most of the methods, such as artificial neural networks, have been focused on estimating hydrological data based on observation over time. Even though these models provide good results, it can be observed that results become unreliable when the training dataset is small or when input data is significantly out of range compared to the training data. Therefore, a new approach is presented, in which artificial neural networks are trained to satisfy physical laws. This is conducted by a novel method called physics-informed neural networks (PINNs), in which physical principles are embedded in a custom loss function. This paper presents the application of physics informed neural networks for solving 1D flood wave propagation in open channels. The research has shown promising results.", journal = "1st Serbian International Conference on Applied Artificial Intelligence (SICAAI)", title = "Physics informed neural networks for 1D flood routing", url = "https://hdl.handle.net/21.15107/rcub_grafar_2771" }
Bojović, F., Milašinović, M., Jovanović, B., Krstić, L., Stojanović, B., Ivanović, M., Prodanović, D.,& Milivojević, N.. (2022). Physics informed neural networks for 1D flood routing. in 1st Serbian International Conference on Applied Artificial Intelligence (SICAAI). https://hdl.handle.net/21.15107/rcub_grafar_2771
Bojović F, Milašinović M, Jovanović B, Krstić L, Stojanović B, Ivanović M, Prodanović D, Milivojević N. Physics informed neural networks for 1D flood routing. in 1st Serbian International Conference on Applied Artificial Intelligence (SICAAI). 2022;. https://hdl.handle.net/21.15107/rcub_grafar_2771 .
Bojović, Filip, Milašinović, Miloš, Jovanović, Branka, Krstić, Lazar, Stojanović, Boban, Ivanović, Miloš, Prodanović, Dušan, Milivojević, Nikola, "Physics informed neural networks for 1D flood routing" in 1st Serbian International Conference on Applied Artificial Intelligence (SICAAI) (2022), https://hdl.handle.net/21.15107/rcub_grafar_2771 .