Reference for citation: Barhoum Z. A., Kolyubin S. A. Physically Consistent Dynamic Modeling of Underwater Robots for Robust Long-Horizon Motion Prediction. Journal of Instrument Engineering. 2025. Vol. 68, N 11. P. 983–995. DOI: 10.17586/0021-3454-2025-68-11-983-995.

Abstract. Modeling the dynamics of underwater robots is a complex task due to the presence of both parametric and functional uncertainties. These arise from interactions with a viscous medium, a priori uncertainty, and variability in the system’s dynamic parameters, as well as the complexity and computational cost of first-principles models and the challenges of identification procedures. This paper proposes the use of neural network parameterization of ordinary differential equations based on the port-Hamiltonian formalism to develop accurate and computationally efficient dynamic models of underwater robots. These models can be used for trajectory prediction, integration with onboard sensor data for localization systems, and controller synthesis. The proposed approach captures both the physical structure of the system and the impact of uncertainties, enabling the creation of physically grounded, data-driven representations of complex nonlinear dynamics. Comparative experiments with classical identification and modeling methods using real-world data from an underwater robot demonstrate advantages of the proposed method in prediction accuracy and its robustness over long-horizon prediction.

Keywords: port-Hamiltonian neural networks, underwater robots, system identification, motion prediction over long time horizons, functional uncertainty, parametric uncertainty, energy-based modeling, neural network parameterization of ordinary differential equations

Acknowledgement: This research was supported by ITMO University Research Projects in AI Initiative (RPAII), no. 640105.

References:

1. Choi H.T., Yuh J. Underwater robots, Springer Handbook of Robotics, 2016, рр. 595–622.
2. Zhu D., Yan T., Yang S.X. arXiv preprint arXiv:2207.04360, 2022.
3. Antonelli G., Fossen T.I., Yoerger D.R. Modeling and control of underwater robots, Springer Handbook of Robotics, 2016, рр. 1285–1306.
4. Zhao Y. et al. Journal of Marine Science and Engineering, 2024, no. 5(12), pp. 801.
5. Ramirez W.A. et al. International Journal of Automation and Computing, 2021, no. 5(18), pp. 681–693.
6. Wu H.M., Karkoub M. Advances in Mechanical Engineering, 2021, no. 10(13), pp. 16878140211053429.
7. Kong F., Guo Y., Lyu W. IEEE Access, 2020, vol. 8, рр. 30119–30126.
8. Valeriano-Medina Y. et al. Mathematical and Computer Modelling of Dynamical Systems, 2013, no. 2(19), pp. 175–200.
9. Javanmard E., Mansoorzadeh S. Journal of Applied Fluid Mechanics, 2019, no. 3(12), pp. 947–959.
10. De Barros E.A., Pascoal A., De Sa E. Ocean Engineering, 2008, no. 16(35), pp. 1627–1636.
11. Faros I., Tanner H.G. arXiv preprint arXiv:2504.02005, 2025.
12. Cortez W.S. et al. IFAC-PapersOnLine, 2023, no. 1(56), pp. 228–233.
13. Sousa C.D., Cortesao R. IEEE/ASME Transactions on Mechatronics, 2019, no. 1(24), pp. 406–411.
14. Sutanto G. et al. Learning for Dynamics and Control, PMLR, 2020, рр. 804–813.
15. Reuss M. et al. arXiv preprint arXiv:2205.13804, 2022.
16. Wang J. et al. Automation and Remote Control, 2018, vol. 79, рр. 2159–2168.
17. Muñoz F. et al. Applied Sciences, 2021, no. 6(11), pp. 2797.
18. Szymkowiak M. Transportation Research Procedia, 2025, vol. 83, рр. 417–424.
19. Singh M., Alexis K. arXiv preprint arXiv:2502.07726, 2025.
20. Sun H. et al. Journal of Marine Science and Engineering, 2022, no. 5(10), pp. 575.
21. Raissi M., Perdikaris P., Karniadakis G.E. arXiv preprint arXiv:1711.10561, 2017.
22. Raissi M., Perdikaris P., Karniadakis G.E. arXiv e-prints, 2017, рр. arXiv: 1711.10566.
23. Lei L., Gang Y., Jing G. Applied Ocean Research, 2022, vol. 121, рр. 103082.
24. Greydanus S., Dzamba M., Yosinski J. Advances in neural information processing systems, 2019, vol. 32.
25. Cranmer M. et al. arXiv preprint arXiv:2003.04630, 2020.
26. Lutter M., Ritter C., Peters J. Deep lagrangian networks: Using physics as model prior for deep learning. arXiv preprint arXiv:1907.04490. 2019.
27. Mirzai B. Physics-informed deep learning for system identification of autonomous underwater vehicles: A Lagrangian neural network approach, 2021.
28. Duong T., Atanasov N. arXiv preprint arXiv:2106.12782, 2021.
29. Duong T. et al. IEEE Transactions on Robotics, 2024.
30. Neary C., Topcu U. Learning for Dynamics and Control Conference, PMLR, 2023, рр. 679–691.
31. Chen R.T.Q. et al. Advances in neural information processing systems, 2018, vol. 31.
32. Singh M., Dharmadhikari M., Alexis K. 2024 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2024, рр. 10005–10011.