多维最大熵分布及其在南海平台防灾设计中的应用研究

Through collecting the topography, hydrologic, meteorological data affecting typhoons and historical records of typhoon disaster on oil fields in the South China Sea, high-accuracy numerical model will be established for typhoon processes. The data observed at ocean observation stations and measured on marine platforms will be utilized to improve the accuracy of the numerical model validation. Time curves of wind, waves, current and tide in typhoon processes across 30 consecutive years are calculated for establishment of database in the South China Sea. Multi-samples combined by extreme environmental factor with concomitant environmental factors will be abstracted. According to the principle of maximum entropy, multivariate maximum entropy model can be constructed. Based on Copula function, considering the random point process of typhoon occurrence, compound multivariate distribution will be proposed. And then estimate joint return parameters of multivariate for marine structure design. Determine ocean environmental condition criteria for platform operation. Obtain wave spectra in different regions the South China Sea for fatigue analysis of offshore structures design. The developed theory and methods will improve and perfect the design specifications. The study results will provide technical support for the structural safety of the South China Sea platform.

调查收集影响南海台风的地形、水文、气象等资料，并收集台风对油田的灾害损失记录，建立南海台风过程后报的数值模式，采用现有海洋观测台站或南海石油平台上实测到的台风个例，对改进的数值模式进行计算精度验证。对连续30年的台风的风、浪、流、潮时程进行重现，建立南海台风过程的海洋环境要素数据库。抽取获得不同环境要素为主极值时的多维样本数据。根据最大熵原则，构建多维最大熵模型。基于Copula函数，考虑台风发生的随机点过程，构建复合型的多维分布。进而估计多种环境要素联合重现值设计参数；确定海洋平台作业环境条件标准；获得南海不同区域海浪谱，用于海洋结构设计的疲劳分析。为设计规范的改进完善提供理论依据，为南海平台结构安全提供技术支撑。

调查收集影响南海台风的地形、水文、气象等资料，建立南海台风过程后报的数值模式，采用现有海洋观测台站或南海石油平台上实测到的台风个例，对改进的数值模式进行计算精度验证。对连续30年的台风的风、浪、流、潮时程进行重现，形成南海台风过程的海洋环境要素数时间历程。抽取获得不同环境要素为主极值时的多维样本数据。根据最大熵原则，构建多维最大熵模型。基于Copula函数，考虑台风发生的极值过程，构建Copula型的多维分布。进而估计多种环境要素联合重现值设计参数；确定浮式海洋平台作业环境条件标准。为设计规范的改进完善提供理论依据，为南海平台结构安全提供技术支撑。