节点文献

设计洪水估算方法的比较研究

Comparative Study on Design Flood Estimating Methods

【作者】 周芬

【导师】 郭生练;

【作者基本信息】 武汉大学 , 水文学及水资源, 2004, 硕士

【摘要】 设计洪水计算是工程水文学的核心内容和重要组成部分。国内外水文学者在设计洪水领域内做了大量的理论研究、方法探索和实际应用工作,取得了丰硕的研究成果,并积累了许多经验。论文结合水利部重大科研项目《水库设计运用专题研究》之一“国内外水库设计洪水理论和防洪调度方法评价”开展研究工作。论文的主要工作和结论包括: (1)在综述和总结国内外有关文献的基础上,分析探讨了洪水频率分析、无资料地区设计洪水和PMP/PMF估计等的研究进展和存在问题。 (2)采用统计试验的方法对P-Ⅲ型分布的多种参数估计方法进行了比较研究。结果表明:就无偏性而言,对于连序系列,数值积分单、双权函数法较好,对于不连序系列,线性矩法较好;在稳健性方面,对于连序和不连序系列,线性矩法都是最好的。线性矩是样本次序统计量的某种线性组合,在估计中该方法不包含样本值的高次方运算,仅是一阶样本矩的计算,所以估计结果受样本中个别点据误差的影响小,估计的偏差小且更稳健。 (3)区域频率分析能充分利用相似流域组的资料来扩充信息量,克服单站样本系列资料短缺的局限,解决无资料地区设计洪水的估算问题。无资料地区设计洪水的区域频率分析包括选择相似流域组、估计指标洪水以及推求区域综合增长曲线和设计流域的洪水频率曲线。以长江中游和汉江下游的部分流域为例,采用该方法推求了无资料地区的设计洪水,结果表明该方法是可行的,值得进一步研究和推广。 (4)综述国内外可能最大暴雨洪水的研究进展,评价其估算方法的特点。通过分析比较国内外PMP/PMF的估算成果,指出了我国研究所存在的问题,并探讨了造成我国PMP/PMF估算成果偏大的原因,提出和建议今后的研究方向。

【Abstract】 Design flood estimation is the key content and important part of engineering hydrology. A great many of theories and methods of design flood estimation have been studied and applied by hydrologists, and lots of achievements and experience were obtained. This paper was part of the key research projects sponsored by the Ministry of Water Resources, "The assessment of reservoir design flood theories and flood control methods at home and broad". The main works and conclusions were summarized as follows:(1) Based on the review and summary of the available literatures and guidelines of design flood all over the world, the research advancements and existent problems for flood frequency analysis, flood estimation in un-gauged area and PMP/PMF estimators were analyzed and discussed.(2) The statistical experiment was applied to compare different parameter estimation methods for P-Ⅲ distribution. The simulation results showed that: for the un-biasness property, numerical integral with single-weighted or double-weighted function performed well in the sequential series and L-moment(LM) was the best one in the series including historical flood respectively; LM was the best method in both series for the robustness property. LM could be expressed in terms of the linear combinations of order statistics and only made operation of one order sampling moments when it was applied to estimate parameters of some sort of distribution. Therefore, LM was little affected by the errors of several sample plots, and the results by the method of LM were more robust than the other ones.(3) Regional flood frequency analysis is a useful approach method that avoids the shortness of at-site sample series and estimates design flood for un-gauged area. Regional flood frequency analysis for un-gauged basins included selecting a pooling-group, estimating the index flood, deriving pooled growth curve and fitting flood frequency curve for tested basin. A case study on some sub-basins of Yangtze River and Han River indicated that this method was practicable and the further research and application was worthwhile.(4) The advancement of PMP/PMF estimation at home and abroad was introduced and analyzed. The results of PMP/PMF in several countries were compared. The reasons why the PMP/PMF estimators in China were larger than those of other countries were discussed. Some suggestions on the future research of PMP/PMF were also pointed out.

  • 【网络出版投稿人】 武汉大学
  • 【网络出版年期】2004年 04期
  • 【分类号】TV122
  • 【被引频次】11
  • 【下载频次】1120
节点文献中: 

本文链接的文献网络图示:

本文的引文网络