| 基于非对称Archimedean Copula的三变量风浪重现水平分析 |
| |
| 引用本文： | 陈子燊,路剑飞,于吉涛.基于非对称Archimedean Copula的三变量风浪重现水平分析[J].海洋通报,2017,36(6):631-637. |
| |
| 作者姓名： | 陈子燊 路剑飞 于吉涛 |
| |
| 作者单位： | 中山大学水资源与环境系，广东广州510275,国土资源部广州海洋地质调查局，广东广州510075,河南理工大学测绘与国土信息工程学院，河南焦作454000 |
| |
| 基金项目： | 国家自然科学基金（41371498） |
| |
| 摘 要： | 采用非对称Archimedean Copula函数与Kendall分布函数分析极端波况下的波高、周期和风速三变量联合概率分布与风险率及其设计分位数,为海岸海洋工程设计和风险评估提供参考依据。以粤东汕尾海域的实测风浪数据为例,使用非对称Gumbel-Hougaard Copula函数计算三变量风浪联合分布的"或"重现期、"且"重现期和二次重现期及其最可能的风浪设计值。主要结论如下:对比不同设计风浪重现期显示,"或"重现期的风险率偏高,"且"重现期的风险率偏低,二次重现期更准确地反映了特定设计频率情况下三变量风浪的风险率;按目前有关规范设计要求的单变量风浪要素设计值已经达到安全标准,按三变量"或"重现期和三变量同频率设计值推算的风浪设计值偏高,以最大可能概率推算的三变量风浪要素的二次重现期设计值可为相关工程安全与风险管理提供新的选择。
|
| 关 键 词： | 非对称Archimedean Copula 风浪风险评估 Kendall 分布函数 二次重现期 |
| 收稿时间： | 2017/3/27 0:00:00 |
| 修稿时间： | 2017/6/13 0:00:00 |
Analysis on return levels of trivariate stormy waves based on asymmetric Archimedean copula function |
| |
| CHEN Zi-shen,LU Jian-fei and YU Ji-tao.Analysis on return levels of trivariate stormy waves based on asymmetric Archimedean copula function[J].Marine Science Bulletin,2017,36(6):631-637. |
| |
| Authors: | CHEN Zi-shen LU Jian-fei and YU Ji-tao |
| |
| Institution: | Department of Water Resource and Environment, Sun Yat-sen University, Guangzhou 510275, China,Guangzhou Marine Geological Survey, Guangzhou 510075, China and School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, China |
| |
| Abstract: | The joint probability distribution for triple variables (i.e., wave height, wave period and wind speed under extreme wave condition) was analyzed by using 3-dimensional asymmetric Archimedean copula functions. And the trivariate stormy wave risk probabilities were further explored based on the Kendall distribution function in order to provide the basis for the coastal marine engineering design and risk assessment. A case is studied by using annual maximum wave heights and the relevant wave periods and wind speeds measured in Shanwei sea waters of eastern Guangdong, the primary return periods and secondary return periods of trivariate stormy wave joint distribution and the most likely design values were computed by using the asymmetric Gumbel-Hougaard copula. The main conclusions of this study can be summarized as follows: Comparing the risk probabilities of trivariate stormy waves among the different design return periods, the ''OR'' joint return period showed higher risk probabilities, and the risk probabilities of ''AND'' joint return period were lower, while the secondary return periods were more accurately depicted the risk probabilities under these specific design frequencies. According to the relevant specifications of the current design requirements, the univariate wave design values have reached safety standards. The estimated storm wave design values with the trivariate ''OR'' return period and triple variables with the same frequencies were obviously higher than expected. The most-likely design realization of the secondary return period can serve as the new selection for coastal marine engineering and risk management. |
| |
| Keywords: | asymmetric Archimedean copulas risk assessment of stormy waves Kendall distribution function secondary return periods |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《海洋通报》浏览原始摘要信息 |
| 点击此处可从《海洋通报》下载免费的PDF全文 |
|
