摘要
采用Copula函数描述多变量的联合概率分布,在此基础上推导了条件最可能组合的计算公式,构建了钱塘江4个站点的洪峰与时段洪量的频率组合的二维联合分布函数,计算了不同设计标准下两者按最可能组合的估计值及其90%的置信区间;同时对同频率组合与条件期望组合的估计结果进行了对比分析。研究表明,当钱塘江水系设计频率小于14%时(对应于8年一遇以上重现期),条件最可能组合计算的设计值介于同频率组合与条件期望组合的设计值之间,同频率组合的设计值最大。
The theory of Copula is used to establish the joint distribution functions for hydrological variables, and then the formula for calculating the conditional most likely composition is derived. The flood peak series and flood volume series of Qiantang River are selected as a case study, the bivariate Copula function is used to analyze the joint distributions of these two variables. The design values with different design criteria are estimated for three kinds of composition modes, i.e. the conditional most likely composition, the identical frequency composition and the conditional expectation composition. The corresponding 90% confidence intervals are also displayed. The results indicate that, for Qiantang River system, the estimated design floods by conditional most likely composition lies between that of by the identical frequency composition and the conditional expectation composition if the design exceedanee frequency is less than 14% （i.e. corresponding to 8-yr return period）, and values estimated by the identical frequency composition are the maximum
出处
《水力发电》
北大核心
2014年第5期13-16,共4页
Water Power
基金
国家自然科学基金项目(51190095
51079039)
关键词
设计洪水
COPULA函数
条件最可能组合
条件期望组合
同频率组合
design flood
Copula function
conditional most likely composition
conditional expectation composition
identical frequency composition
