Application of statistical methods for the prediction of extreme wave events / Barbara Paplinska-Swerpel in Journal of hydraulic research, Vol. 46 extra issue (Mars/Avril 2008)  

Public ISBD [article]  inJournal of hydraulic research > Vol. 46 extra issue (Mars/Avril 2008) . - pp.314-323 Titre : Application of statistical methods for the prediction of extreme wave events Titre original : Application de méthodes statistiques à la prédiction d'événements de mer extrêmes Type de document : texte imprimé Auteurs : Barbara Paplinska-Swerpel, Auteur ; Lukasz Paszke, Auteur ; Wojciech Sulisz, Auteur Article en page(s) : pp.314-323 Note générale : Hydraulique Résumé en Français Langues : Anglais (eng) Mots-clés : Extreme sea states forecast Neural networks Sea waves modeling Index. décimale : 627 Ingénierie des cours d'eau naturels, des ports, des rades et des cotes. Installations de navigation, de dragage, de récupération et de sauvetage. Barrages et centrales électriques hydrauliques Résumé : A statistically-based model is applied to forecast sea states for severe storms. The model is based on the application of a neural network and predicts extreme sea state parameters at specified locations. The results show that the neural network can be applied to forecast extreme sea state parameters. This requires a special treatment of the input data. The analysis shows that different types of input data and training data sets should be considered and the representativity of the training data set must be improved. Moreover, a sensitivity analysis should be conducted to remove excess information from the input data. The processing of data sets significantly reduces the number of parameters applied in the model and improves the prediction for most severe storms. The analysis indicates that this neural network model may be helpful in the selection of a measurement system for the forecasting of extreme sea state parameters. This is important because typical installations of wave buoys along the coast have a limited forecasting applicability range. DEWEY : 627 ISSN : 0022-1686 En ligne : http://www.journalhydraulicresearch.com [article] Application of statistical methods for the prediction of extreme wave events = Application de méthodes statistiques à la prédiction d'événements de mer extrêmes [texte imprimé] / Barbara Paplinska-Swerpel, Auteur ; Lukasz Paszke, Auteur ; Wojciech Sulisz, Auteur . - pp.314-323. Hydraulique Résumé en Français Langues : Anglais (eng) in Journal of hydraulic research > Vol. 46 extra issue (Mars/Avril 2008) . - pp.314-323 Mots-clés : Extreme sea states forecast Neural networks Sea waves modeling Index. décimale : 627 Ingénierie des cours d'eau naturels, des ports, des rades et des cotes. Installations de navigation, de dragage, de récupération et de sauvetage. Barrages et centrales électriques hydrauliques Résumé : A statistically-based model is applied to forecast sea states for severe storms. The model is based on the application of a neural network and predicts extreme sea state parameters at specified locations. The results show that the neural network can be applied to forecast extreme sea state parameters. This requires a special treatment of the input data. The analysis shows that different types of input data and training data sets should be considered and the representativity of the training data set must be improved. Moreover, a sensitivity analysis should be conducted to remove excess information from the input data. The processing of data sets significantly reduces the number of parameters applied in the model and improves the prediction for most severe storms. The analysis indicates that this neural network model may be helpful in the selection of a measurement system for the forecasting of extreme sea state parameters. This is important because typical installations of wave buoys along the coast have a limited forecasting applicability range. DEWEY : 627 ISSN : 0022-1686 En ligne : http://www.journalhydraulicresearch.com

Exemplaires

Code-barres	Cote	Support	Localisation	Section	Disponibilité
aucun exemplaire