TY - GEN
T1 - Physics-Informed Echo State Networks for Chaotic Systems Forecasting
AU - Doan, Nguyen Anh Khoa
AU - Polifke, Wolfgang
AU - Magri, Luca
PY - 2019
Y1 - 2019
N2 - We propose a physics-informed Echo State Network (ESN) to predict the evolution of chaotic systems. Compared to conventional ESNs, the physics-informed ESNs are trained to solve supervised learning tasks while ensuring that their predictions do not violate physical laws. This is achieved by introducing an additional loss function during the training of the ESNs, which penalizes non-physical predictions without the need of any additional training data. This approach is demonstrated on a chaotic Lorenz system, where the physics-informed ESNs improve the predictability horizon by about two Lyapunov times as compared to conventional ESNs. The proposed framework shows the potential of using machine learning combined with prior physical knowledge to improve the time-accurate prediction of chaotic dynamical systems.
AB - We propose a physics-informed Echo State Network (ESN) to predict the evolution of chaotic systems. Compared to conventional ESNs, the physics-informed ESNs are trained to solve supervised learning tasks while ensuring that their predictions do not violate physical laws. This is achieved by introducing an additional loss function during the training of the ESNs, which penalizes non-physical predictions without the need of any additional training data. This approach is demonstrated on a chaotic Lorenz system, where the physics-informed ESNs improve the predictability horizon by about two Lyapunov times as compared to conventional ESNs. The proposed framework shows the potential of using machine learning combined with prior physical knowledge to improve the time-accurate prediction of chaotic dynamical systems.
KW - Chaotic dynamical systems
KW - Echo State Networks
KW - Physics-Informed Neural Networks
UR - http://www.scopus.com/inward/record.url?scp=85067603854&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-22747-0_15
DO - 10.1007/978-3-030-22747-0_15
M3 - Conference contribution
AN - SCOPUS:85067603854
SN - 9783030227463
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 192
EP - 198
BT - Computational Science – ICCS 2019 - 19th International Conference, Proceedings
A2 - Rodrigues, João M.F.
A2 - Cardoso, Pedro J.S.
A2 - Monteiro, Jânio
A2 - Lam, Roberto
A2 - Krzhizhanovskaya, Valeria V.
A2 - Lees, Michael H.
A2 - Sloot, Peter M.A.
A2 - Dongarra, Jack J.
PB - Springer
T2 - 19th International Conference on Computational Science, ICCS 2019
Y2 - 12 June 2019 through 14 June 2019
ER -