- Research Article
1
- 10.1080/09377255.2025.2509378
A machine learning algorithm to estimate hydrodynamic coefficients of a jack-up vessel in frequency domain
- Jun 20, 2025
- Ship Technology Research
- Per Mikkelsen
ABSTRACT This paper presents the development of a machine learning (ML) algorithm to predict the hydrodynamic coefficients to calculate the motion behavior of a vessel in frequency domain. The aim is to develop a less laborious and faster method compared to the conventional procedure for determining the hydrodynamic coefficients for added mass, damping, and excitation forces by radiation-diffraction models with linear potential flow solvers based on Boundary Element Methods (BEM). Physics-informed machine learning was utilized to incorporate physical laws directly into the ML models, ensuring accurate and physically consistent predictions. The ML-algorithm is trained and validated with a database created by the software ANSYS AQWA. The approach is performed on the Jack-Up Vessel(JUV) Innovation, owned and operated by DEME.
Read more