Kriging

Predicting Ship Interactions: A Data-Driven Approach to Safer Berthing

Non-parallel ship interactions generate strongly asymmetric hydrodynamic forces that standard models struggle to predict. Our new study builds 18 Kriging models from unsteady RANS, CFD data, testing how berthing angle, water depth, and lateral spacing affect surge, sway, and yaw on KVLCC2 and Aframax hulls. The results are instructive: reliability is response-specific. CaseA achieved held-out R² values of 0.96-1.00, while other groups showed narrower predictive power. This is not a one-size-fits-all solution, it is a calibrated, validated step toward safer berthing.