Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
A generalized probabilistic model of ice load peaks on ship hulls in broken-ice fields
https://orcid.org/0000-0002-8964-9812
Abstract On some occasions, especially when traveling in a broken ice field, the process of ice induced loads on ship hulls cannot be well modeled by traditional statistical models which are usually applied for this purpose. Examples of such are the one-parameter exponential and the Weibull's models. The fitting achieved by application of these models could generally result in underestimation of the predicted extreme values. Therefore a so-called generalized model, i.e. the three-parameter exponential model, is proposed in order to improve the process of fitting. The proposed model is actually a proportional combination of two one-parameter exponential models. It tends to give more conservative predicted extreme values as compared to the one-parameter exponential and the Weibull's models. Various approaches for estimation of the parameters are treated, i.e. the method of moments, the non-linear least square method, the non-linear least square method based on application of Kernel density estimation, and the maximum likelihood estimators. The fitting by means of the non-linear least square method was observed to give the best results. However, more stable predictions are provided by the maximum likelihood estimators.
Highlights Offers the three-parameter exponential as a generalized statistical ice load model Provides the derivation of the properties of the three-parameter exponential model Gives parameterization of the non-parametric model based on Kernel density estimation Performs comparison between various statistical models and fitting methods
A generalized probabilistic model of ice load peaks on ship hulls in broken-ice fields
https://orcid.org/0000-0002-8964-9812
Abstract On some occasions, especially when traveling in a broken ice field, the process of ice induced loads on ship hulls cannot be well modeled by traditional statistical models which are usually applied for this purpose. Examples of such are the one-parameter exponential and the Weibull's models. The fitting achieved by application of these models could generally result in underestimation of the predicted extreme values. Therefore a so-called generalized model, i.e. the three-parameter exponential model, is proposed in order to improve the process of fitting. The proposed model is actually a proportional combination of two one-parameter exponential models. It tends to give more conservative predicted extreme values as compared to the one-parameter exponential and the Weibull's models. Various approaches for estimation of the parameters are treated, i.e. the method of moments, the non-linear least square method, the non-linear least square method based on application of Kernel density estimation, and the maximum likelihood estimators. The fitting by means of the non-linear least square method was observed to give the best results. However, more stable predictions are provided by the maximum likelihood estimators.
Highlights Offers the three-parameter exponential as a generalized statistical ice load model Provides the derivation of the properties of the three-parameter exponential model Gives parameterization of the non-parametric model based on Kernel density estimation Performs comparison between various statistical models and fitting methods
A generalized probabilistic model of ice load peaks on ship hulls in broken-ice fields
https://orcid.org/0000-0002-8964-9812
Abstract On some occasions, especially when traveling in a broken ice field, the process of ice induced loads on ship hulls cannot be well modeled by traditional statistical models which are usually applied for this purpose. Examples of such are the one-parameter exponential and the Weibull's models. The fitting achieved by application of these models could generally result in underestimation of the predicted extreme values. Therefore a so-called generalized model, i.e. the three-parameter exponential model, is proposed in order to improve the process of fitting. The proposed model is actually a proportional combination of two one-parameter exponential models. It tends to give more conservative predicted extreme values as compared to the one-parameter exponential and the Weibull's models. Various approaches for estimation of the parameters are treated, i.e. the method of moments, the non-linear least square method, the non-linear least square method based on application of Kernel density estimation, and the maximum likelihood estimators. The fitting by means of the non-linear least square method was observed to give the best results. However, more stable predictions are provided by the maximum likelihood estimators.
Highlights Offers the three-parameter exponential as a generalized statistical ice load model Provides the derivation of the properties of the three-parameter exponential model Gives parameterization of the non-parametric model based on Kernel density estimation Performs comparison between various statistical models and fitting methods
A generalized probabilistic model of ice load peaks on ship hulls in broken-ice fields
Suyuthi, A. (Autor:in) / Leira, B.J. (Autor:in) / Riska, K. (Autor:in)
Cold Regions, Science and Technology ; 97 ; 7-20
26.09.2013
14 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
A generalized probabilistic model of ice load peaks on ship hulls in broken-ice fields
| Online Contents | 2014
Statistics of local ice load peaks on ship hulls
| British Library Online Contents | 2013
Statistics of local ice load peaks on ship hulls
| Elsevier | 2012
Statistics of local ice load peaks on ship hulls
| Online Contents | 2013
Collapase analysis of ship hulls
| Elsevier | 1993