A geometry-aware per-surface neural network surrogate predicts real-time hydrodynamic forces for amphibious AGVs from SDF submergence data, achieving low sMAPE on held-out CFD and reproducing quadratic drag and linear buoyancy scaling in full-scale vehicle trials without explicit encoding in theloss
Transient, 3d cfd, moving mesh simulation of vehicle water wading in a water tunnel with inclined entry-exit,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Geometry-Aware Surrogate for Real-Time Hydrodynamics Estimation of Autonomous Ground Vehicles in Amphibious Environments
A geometry-aware per-surface neural network surrogate predicts real-time hydrodynamic forces for amphibious AGVs from SDF submergence data, achieving low sMAPE on held-out CFD and reproducing quadratic drag and linear buoyancy scaling in full-scale vehicle trials without explicit encoding in theloss