A NEURAL NETWORK APPROACH FOR ESTIMATING THE METALLIC HULL WEIGHT OF TRANSPORT SHIPS

This paper deals with the neural estimation of the metallic hull weight for transport ships based on the multi-layer feedforward neural network model trained by using the backpropagation learning algorithm. It is shown by the computation results for bulk carriers and oil tankers that massively parallel, interconnected networks of nonlinear analog neurons are viable and effective in the hull weight estimation.