论文部分内容阅读
本文介绍了以两种不同类型的样本求取多年一遇极值的推算方法。第一种类型是用长期资料的年极值为样本,分别依耿贝尔极值分市(矩法估算参数和最小二乘法估算参数)、第一极值分布、皮尔逊Ⅲ型分布,推算了多年一遇的风暴增水和水位;对所有这些分布的推算结果,用柯尔莫哥洛夫方法进行了测适度检验,给出了95%的可信度。第二种类型用短期连续资料力样本,以三个月的实测水位资料用Weibull分布推算了不同重现期的水位。用两种类型推算的结果是相同的。
This article describes how to find the extreme value of one year in two different types of samples. The first type is based on the annual extremum of long-term data, which are respectively calculated according to the Gunge-Bell extreme points market (moment estimation parameters and least squares estimation parameters), the first extreme value distribution and the Pearson Ⅲ type distribution The storm surge and the water level at many years of the year; all these distributions were estimated by means of the Kolmogorov method, giving a 95% confidence level. The second type uses the short-term continuous material force samples to calculate the water level at different recurrence periods with the Weibull distribution based on the three-month measured water level data. The results of the two types of calculations are the same.