Prediction of Surface Ship's Residual Resistance Coefficient Using Neural Networks

Abstract

The Holtrop method, which provides a prediction of the components of surface ships total resistance, is widely used at ships initial design stage for estimating the resistance. In this paper a neural network model which performs the same role as the Holtrop method is presented to predict the residual resistance. A multilayer perceptron has been trained with the data generated by the Holtrop method to learn the relationship between the input (length-displacement ratio, prismatic coefficient, breadth-draft ratio and Froude number) and the target variable (the residual resistance coefficient). The results of this model have been compared against those provided by the Holtrop method and it is found that the quality of the prediction is improved over the entire range of data. The neural network provides an accurate estimation of the residual resistance with the Froude number and the hull geometry coefficients as variables.