A physics-informed neural network with control (PINC) trained on synthetic BlueROV2 trajectories predicts longer-horizon states more accurately than a non-physics-informed network in simulation.
Unav-sim: A visually realistic underwater robotics simulator and synthetic data-generation framework,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Modelling of Underwater Vehicles using Physics-Informed Neural Networks with Control
A physics-informed neural network with control (PINC) trained on synthetic BlueROV2 trajectories predicts longer-horizon states more accurately than a non-physics-informed network in simulation.