基于Nataf变换的相关随机数算法及其应用

Correlative random number algorithm method and its application based on Nataf transformation

针对极值风浪荷载之间的相关性,提出了一种基于Nataf变换的相关非高斯分布随机数算法,用来生成极值风浪荷载随机向量;运用该方法讨论了基于独立风浪的荷载效应和基于相关风浪的荷载效应的极值之间的关系,并且对不同的风浪联合荷载效应的拟合方法进行了评定;应用最小二乘法及极值Ⅰ型分布拟合实际风、浪荷载效应,然后对此分布模型运用Nataf变换随机数算法和一般随机数算法生成随机向量,最后通过与实际风浪联合荷载效应比较,验证了所提方法的有效性。

With the correlation between extreme wind and wave-loads, a random number algorithm method of the correlative non-Gaussian distribution based on Natal transformation was proposed to generate random vectors of extreme wind and wave loads. This method was then applied to study the correlation of extreme values between mutu- ally independent load effects and correlative load effects and evaluate the different fitting methods for wind and wave load effects. Actual wind and wave loads effects were fitted using the least square method and extreme value type I distribution, random vectors were generated by Nataf transformation random number algorithm method and general random number algorithm method using this distribution, and then, the validity of the method was confirmed through comparing the generated random vectors with practical data.