Abstract

Physics-informed neural networks (PINNs) integrate physical laws withdata-driven models to improve generalization and sample efficiency. This workintroduces an open-source implementation of the Physics-Informed Neural Networkwith Control (PINC) framework, designed to model the dynamics of an underwatervehicle. Using initial states, control actions, and time inputs, PINC extendsPINNs 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 simulatedunderwater vehicle demonstrates more accurate long-horizon predictions comparedto a non-physics-informed baseline

 

Quick Read (beta)