极值风速拟合优化策略  

Optimal Policy of Extreme Wind Fitting

作　　者：赵林[1] 葛耀君[1] 项海帆[1] 

机构地区：[1]同济大学土木工程防灾国家重点实验室,上海200092

出　　处：《同济大学学报（自然科学版）》2003年第4期383-388,共6页Journal of Tongji University:Natural Science

基　　金：国家自然科学基金资助项目 (5 960 80 0 6);高等学校博士学科点专项科研基金资助项目 (960 2 4713)

摘　　要：在桥梁设计规范中假定风速随机过程为平稳过程 ,风速母体服从指数型分布并以极值Ⅰ型拟合风速极值渐近分布 .然而实际上风速并不严格地满足平稳过程假设 ,且由于样本数据的来源不同和极值分布参数估计方法各异 ,对风速母体分布及样本极值渐近分布拟合方法的结论不尽相同 .针对上述问题 ,通过蒙特卡罗数值模拟技术研究了多种风速母体分布下具有普遍适用性的极值风速拟合策略 .并以上海地区龙华、川沙气象站极值风速估算为例 。The bridge design code presumes that wind velocity stochastic process is a stationary gauss process,and extreme value distribution I is used to fit the extreme wind velocity.However,since sampling data,estimation methods of distribution parameter,wind velocity generating distributions and fitting approaches to extreme value asymptotic distribution,are distinctly employed,various conclusions have been reached about wind velocity generating distributions and fitting approaches of extreme value asymptotic distributions.In order to resolve the above problems,a generally applicable fitting policy under various generating samples of wind velocity distributions is proposed,and its applicability and reasonability are illustrated with the extreme value wind velocity evaluations of Longhua and Chuansha weather stations in Shanghai.

关 键 词：极值风速 极值渐近分布 参数估计 母体分布 蒙特卡罗数值模拟 

分 类 号：TU973.32[建筑科学—结构工程]