Modeling of Vehicle Mobility in Shallow Water with Data-Driven Hydrodynamics Model
Abstract
In this study, a numerical procedure for predicting vehicle mobility in shallow water is proposed with the data-driven hydrodynamics model, considering the effect of soil deformation. To this end, the high-fidelity coupled CFD-MBD model is developed to characterize the hydrodynamic loads exerted on the vehicle in shallow water and used to generate the training dataset for the proposed data-driven model. To account for the history-dependent hydrodynamic force and moment characteristics, LSTM is introduced to account for the effect of the historic alvariation of the vehicle motion states as the input to the data-driven model. The data-driven models are called from the MBD mobility solver at every time step to determine the hydrodynamic loads, allowing for predicting the transient responses of the vehicle interacting with shallow water on deformable soil. It is demonstrated by several numerical examples that the complex vehicle-water interaction behavior was predicted accurately by the proposed data-driven hydrodynamics model while achieving a substantial computational speedup. The predictive ability and computational benefit of the proposed hybrid LSTM-MBD vehicle-water interaction model are demonstrated.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 10, 2023
- Accession Number
- AD1218371
Entities
People
- Arkady Grunin
- Hiroki Yamashita
- Hiroyuki Sugiyama
- J. Ezequiel MartÃn
- Nathan Tison
- Paramsothy Jayakumar