Ship Manoeuvrability Prediction Using Neural Networks Analysis
Remus Zăgan, Mihaela-Greti Manea, Emil Manea · Advanced materials research · 2014
Ship motion forecasting is very important for safety of ships especially when operating in offshore mooring state. It is known that the ship motions have dynamical and nonlinear characteristics in the ocean and sea environments. In our paper we try to predict the manoeuvrability of the ships applying the predicted nonlinear wave field with the current state of the vessel motions using ship course time series prediction, which is based on back propagation neural network structure and algorithm, was proposed. The results of simulations performed by means of the elaborated networks are given in comparison with test simulations cases for different value of rudder angle. This method was applied to ship manoeuvrability prediction, and simulation results showed the validity to improving the prediction accuracy.