Ship Residuary Resistance Prediction with Machine Learning

This paper explores machine learning techniques as cost-effective, accurate, and relatively fast alternative to traditional methods for yacht residuary resistance calculation. Currently, solving CFD equations is a well-known method in residuary resistance calculation. However, solving CFD equations requires an iterative process which is very time consuming and needs high performance processing units in addition. Instead of solving CFD equation for any ship geometry, we proposed to apply machine learning on present data obtained from previous CFD-based calculation or measurements of real models. We used Delft yacht hydrodynamics dataset obtained from ۳۰۸ experiments in calculating residuary resistance corresponding to different ship parameters. This dataset is divided into training and test data and the results of several learning-based method are compared in terms of RMSE. Our results demonstrate that XGBoost regressor is superior that others with RMSE=۰.۵۴.

Keywords:

ship residuary resistance , machine learning , Delft dataset , regression