Advanced Machine Learning for Ship Resistance Prediction

Key Points

• The aim is to improve predictions of ship resistance components using a combination of machine learning and physics-based methods.
• Integrated machine learning with traditional physics-based approaches.
• Predicted resistance components such as frictional and wave resistance.
• Maintained interpretability while modeling complex interactions across different hull shapes.
• Achieved improved prediction accuracy of resistance components compared to traditional methods.
• Enhanced understanding of nonlinear interactions in ship resistance.
• Retained physical interpretability in model outputs.

Abstract