| 摘要: |
| 【目的】研究低删失样本部分概率权重矩在洪水频率分布参数估计中的应用。【方法】选取广义极值分布为总体分布线型，利用MATLAB编程实现参数数值求解，采用统计试验研究估计量统计特性，并以陕北地区主要水文站年最大洪峰流量系列为例进行频率分析。【结果】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，在4种拟合效果评价标准下，删失样本下界F0=0.2时理论分布对大多数测站的洪水资料有较好的拟合效果。【结论】部分概率权重矩法可作为陕北地区洪水频率分布参数的估计方法，并可应用于以推求大重现期设计洪水为目的的洪水频率分析。 |
| 关键词: 洪水预测 频率分布 参数估计 部分概率权重矩 广义极值分布 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目（51179160，50879070，50579065）；高等学校博士学科点专项科研基金项目（20110204110017） |
|
| Application of partial probability weighted moments for estimating parameters of flood frequency distribution |
|
|
| Abstract: |
| 【Objective】This study investigated the parameter estimation of flood frequency distribution based on partial probability weighted moments (PPWM).【Method】MATLAB programs were used to achieve-numerical solution of parameters of the generalized extreme value distribution (GEV) from PPWM.Monte Carlo experiments of PPWM were performed to study the statistic properties of estimates.And the application was made to estimating parameters of annual maximum floods peak flow series of main gauging stations in northern Shaanxi.【Result】PPWM method had good effectiveness in high quantile estimation. Using four goodness-of-fit assessments,the theoretical distributions fit the majority of flood data in northern Shaanxi well when the value of lower censored bound F0 was 0.2.【Conclusion】PPWM method can be used for estimating parameters of flood frequency distribution in northern Shaanxi,and in flood frequency analysis aiming to get the design floods that have long return period. |
| Key words: flood forecast frequency distribution parameter estimating partial probability weighted moments generalized extreme value distribution |
