An ANN based system for forecasting ship roll motion

Author

Lopez Pena, F. ; Miguez Gonzalez, M. ; Diaz Casas, V. ; Duro, R.J. ; Pena Agras, D.

Author_Institution

Integrated Group for Eng. Res., Univ. of A Coruna, Ferrol, Spain

fYear

2013

fDate

15-17 July 2013

Firstpage

168

Lastpage

173

Abstract

An ANN based system has been developed for forecasting the roll motion of a ship and predicting the onset of parametric roll resonance. This kind of instability can be devastating for the ship and is a phenomenon that is difficult to predict when using classical mathematical modeling approaches. In the present investigation the ANNs are trained using data obtained from a mathematical model of ship roll motion while the performance of the whole system is verified with realistic towing tank tests. The results achieved are quite promising and support the claim that it can be implemented in any ship without the need for any kind of water tank or real ship tests.

Keywords

hydrodynamics; learning (artificial intelligence); mechanical engineering computing; mechanical stability; neural nets; resonance; ships; tanks (containers); vehicle dynamics; ANN training; ANN-based system; mathematical model; parametric roll resonance onset prediction; ship instability; ship roll motion forecasting; towing tank tests; Artificial neural networks; Forecasting; Marine vehicles; Mathematical model; Testing; Time series analysis; Training; ANN; forecasting; parametric roll;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA), 2013 IEEE International Conference on

Conference_Location

Milan

Print_ISBN

978-1-4673-4701-3

Type

conf

DOI

10.1109/CIVEMSA.2013.6617415

Filename

6617415