标题:
基于部分概率权重矩的洪水频率参数估计方法Estimation of GEV Distribution Parameters Using Partial Probability Weighted Moments
作者:
原秀红, 宋松柏
关键字:
部分概率权重矩, 广义极值分布, Monte Carlo试验, 参数估计Partial Probability Weighted Moments; Generalized Extreme Value Distribution; Monte Carlo Experiments; Parameter Estimating
期刊名称:
《Journal of Water Resources Research》, Vol.1 No.5, 2012-10-31
摘要:
在洪水频率分析中,为了避免小洪水估算大重现期设计洪水值呈现出的滋扰行为,本文应用部分概率权重矩进行洪水频率分布参数估计的原理和方法,探索部分概率权重矩(PPWM)在广义极值分布(GEV)参数估计中的应用。采用MATLAB编程进行基于部分概率权重矩的广义极值分布参数数值求解。通过Monte Carlo试验,研究了不同低删失样本的部分概率权重矩法估计量的统计特性,并获得了相应的统计试验结果。结果表明,低删失样本的部分概率权重矩法在高分位数估计方面呈现出良好的有效性,可以应用于以推求大重现期设计洪水为目的的洪水频率分析。
In order to reduce the uncertainty of the design flood estimation with larger return period, the prin-ciples and methods of Partial probability weighted moments (PPWM) were used to estimate parameters of Generalized extreme distribution (GEV). MATLAB programs were used to achieve the parameters’ numerical solution of the GEV from PPWM on the basis of previous studies. Monte Carlo experiments were performed to assess the statistical properties of parameters estimation of the GEV distribution by the method of PPWM from different lower bound censored samples. The research results indicate that the method of PPWM from lower bound censored samples still has good effectiveness in high quantile estimation.