Score: 1

Modelling of Underwater Vehicles using Physics-Informed Neural Networks with Control

Published: April 28, 2025 | arXiv ID: 2504.20019v1

By: Abdelhakim Amer , David Felsager , Yury Brodskiy and more

Potential Business Impact:

Helps underwater robots learn to move better.

Business Areas:
Autonomous Vehicles Transportation

Physics-informed neural networks (PINNs) integrate physical laws with data-driven models to improve generalization and sample efficiency. This work introduces an open-source implementation of the Physics-Informed Neural Network with Control (PINC) framework, designed to model the dynamics of an underwater vehicle. Using initial states, control actions, and time inputs, PINC extends PINNs to enable physically consistent transitions beyond the training domain. Various PINC configurations are tested, including differing loss functions, gradient-weighting schemes, and hyperparameters. Validation on a simulated underwater vehicle demonstrates more accurate long-horizon predictions compared to a non-physics-informed baseline

Country of Origin
🇩🇰 Denmark

Repos / Data Links

Page Count
8 pages

Category
Computer Science:
Machine Learning (CS)