Ship roll motion time series forecasting using neural networks

Author

Peña, Fernando Lopez ; Gonzalez, Marcos Miguez ; Casás, Vicente Díaz ; Duro, Richard J.

Author_Institution

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

fYear

2011

fDate

19-21 Sept. 2011

Firstpage

1

Lastpage

6

Abstract

A neural network based system has been applied for forecasting the large amplitude roll motions of a ship that appear during parametric roll resonance. Under these conditions, ship roll motion presents a highly nonlinear behavior and accurate predictions are difficult to achieve using classical mathematical modeling approaches. The results obtained present very good agreement to reality, leading to the possibility of applying the system as a base for a parametric roll warning system.

Keywords

alarm systems; forecasting theory; goods distribution; neural nets; ships; time series; amplitude roll motion; classical mathematical modeling approach; neural network based system; nonlinear behavior; parametric roll resonance; parametric roll warning system; ship roll motion time series forecasting; Artificial neural networks; Marine vehicles; Mathematical model; Neurons; Predictive models; Time frequency analysis; Time series analysis; ANN; forecasting; parametric roll;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence for Measurement Systems and Applications (CIMSA), 2011 IEEE International Conference on

Conference_Location

Ottawa, ON, Canada

ISSN

2159-1547

Print_ISBN

978-1-61284-924-9

Type

conf

DOI

10.1109/CIMSA.2011.6059920

Filename

6059920