Sample Selection Method for Ship Resistance Performance Optimization Based on Approximated Model

Abstract

Approximated model instead of computational fluid dynamics tool is utilized for performance analysis in hull form optimization process, which can save time significantly. Sample selection is the central issue of approximated model building. This article focuses on the sample selection method and the application of approximated model in hull form optimization. Latin hypercube sampling and uniform design are compared. The uniform design based on genetic algorithm approach is proposed. The radial basis function interpolation method is used for hull surface automatic modification. An approximated model using a neural network for ship resistance performance is established. The stem profile's optimization for the Korea Research Institute of Ships and Ocean Engineering (KRISO) container ship is completed. The results show that hull form optimization based on the approximated model can significantly improve optimization efficiency and is practical for engineering design.