Floating-Body Hydrodynamic Neural Networks
This addresses the challenge of interpretable and stable modeling of dissipative dynamics in engineering and natural systems, representing a novel method rather than an incremental improvement.
The paper tackled the problem of modeling dissipative fluid-structure interactions in floating-body motion, which is difficult with black-box neural models due to limited interpretability and unstable predictions. The result was the proposed FHNN framework, which achieved up to an order-of-magnitude lower error than Neural ODEs on synthetic vortex datasets and recovered physically consistent flow fields.
Fluid-structure interaction is common in engineering and natural systems, where floating-body motion is governed by added mass, drag, and background flows. Modeling these dissipative dynamics is difficult: black-box neural models regress state derivatives with limited interpretability and unstable long-horizon predictions. We propose Floating-Body Hydrodynamic Neural Networks (FHNN), a physics-structured framework that predicts interpretable hydrodynamic parameters such as directional added masses, drag coefficients, and a streamfunction-based flow, and couples them with analytic equations of motion. This design constrains the hypothesis space, enhances interpretability, and stabilizes integration. On synthetic vortex datasets, FHNN achieves up to an order-of-magnitude lower error than Neural ODEs, recovers physically consistent flow fields. Compared with Hamiltonian and Lagrangian neural networks, FHNN more effectively handles dissipative dynamics while preserving interpretability, which bridges the gap between black-box learning and transparent system identification.