|
|
Publications
Moreira, L. and Guedes Soares, C. (2003), “Training Recurrent Neural Networks with Noisy Data for Manoeuvring Simulation”, Proceedings of the 2nd International Conference on Computer Applications and Information Technology in the Maritime Industries (COMPIT'03), 14-17 May, Hamburg, Germany, pp.183-195
A Recursive Neural Network (RNN) manoeuvring simulation model for surface ships is presented. Inputs to the simulation are the orders of rudder angle and ship’s speed and also the recursive outputs velocities of sway and yaw. The model is used to test the capabilities of artificial neural networks (ANNs) in manoeuvring simulation of ships when the data used for training is noisy. The simulations will be performed using as basis the ship Mariner. The data generated to train the network is obtained through a manoeuvrability model implemented in a block diagram, performing the simulation of different manoeuvring tests. The noisy data is generated introducing white noise into two inputs of the system: the rudder angle and advance speed of the ship. The results achieved using RNNs and the conventional mathematical model are compared in order to analyse the accuracy of the RNN and its generalization ability.
If you did not manage to obtain a copy of this paper: Request a copy of this article
For information about all CENTEC publications you can download: Download the Complete List of CENTEC Publications
|