摘要
指数Γ分布有四个独立参数。分布曲线的弹性好,适应性强,与我国各地区洪水系列点据的配合较好。但是,由于它有四个参数,使得分布参数的估计比较困难。由我国宋德敦、丁晶(1988年)将美国Greenwood等人(1979年)提出的概率权重矩法(PWM),应用于P-Ⅲ分布,取得较好效果。本文将PWM法应用于指数Γ分布的参数估计。统计试验结果和实例分析表明,PWM法用于指数Γ分布的参数估计时,具有较好的统计特性。
Probability weighted moments (PWM) were advanced in 1979 by Greenwood et al. The results in previous research show that the method has better statistical property. In 1988, Song Dedun and Ding Jing further verified that the distributions whose inverse forms must be explicitly defined could by no means be a prerequisite for using the PWM method, and they applied PWM method successfully to the estimation of parameters in Pearson Type-Ill distribution (P-III distribution) although this sort of distribution never has explicit inverse form. In this paper PWM method is employed for the estimation of parameters in Exponential Gamma distribution (Ex-T distribution).The results from Monte Carlo experiments have shown that PWM method is of more favorable statistical characteristics as compared with conventional moments method.
出处
《水利学报》
EI
CSCD
北大核心
1989年第11期1-9,共9页
Journal of Hydraulic Engineering
