【作者】 杨智硕；

【导师】 朱颖元；

【作者基本信息】 福州大学 ， 水工结构工程， 2005， 硕士

【摘要】 近年来，我国社会经济迅速发展，大大促进了城市化建设的步伐，相应地，对城市的给水排水设计也提出了更高的要求。城市水文学作为水文学的分支越来越受到人们的重视。因此，城市排水中设计暴雨的研究成为城市防灾减灾的重要课题之一。 国内外城市水文学的研究已经取得了很大的进展，为城市的排水设计提供了更加科学的依据。我国较为系统地研究城市暴雨及城市排水暴雨强度公式始于70年末期和80年代，其主要内容为暴雨选样方法、短历时暴雨公式参数优选方法，并提出部分城市暴雨强度公式(包括我省福州市、厦门市)。这些研究为我国城市排水减灾、防洪减灾、提高城市环境质量起着重要的作用。但是，城市排水短历时设计暴雨、选样方法采用的是年多个样值。由于自八十年代起，水文部门仅统计年最大值，不再统计年多个样值，使得年多个样值资料很少或者没有，无法满足工程设计的要求。这些都是在实际工程设计中亟待解决的问题。 暴雨选样在城市暴雨公式中早年用超定量法，六十年代起采用年多个样法，并写入历次排水规范中。随着时间的推移，各地暴雨实测资料不断增加。为年最大值的应用提供条件。年最大值的资料容易获得，可以减少工作量，提高工作效率。近几年来，很多学者对年最大值法进行研究，取得许多成果。理论分析表明：年最大值法应用到排水设计是可行，但迄今为止我国还没有进行系统的研究。同时，城市短历时暴雨具有地域性，不同地区有不同的特点。由于上述情况，本文对福建省9个市49座城镇雨量观测站实测的资料进行分析计算，寻找合理的年多个值与年最大值之间的频率转换模型。为制定福建省排水设计暴雨强度公式做准备。其中，选样方式采用年最大值法、年超大值法、超定量法和年多个样法；频率分布模型采用ρ-Ⅲ型分布、耿贝尔分布、指数分布曲线等。为了探讨利用年最大值法代替年多个样法的可能性，利用概率论的原理，建立了四种数学模型，将年重现期(1、2、3、4、5、10、15、20a)转换成次频率，在各自的频率曲线上求出各历时(5、10、15、20、30、45、60、90、120min)的暴雨设计值．并以离差绝对值和、离差平方和以及Nash-Sutcliffe作为误差评定标准，分别计算各自的暴雨设计值的离差绝对值和、离差平方和和NSE值。最后分别计算49个站点不同历时离差绝对值和平均值、离差平方和平均值以及NSE值平均值。计算结果表明：频率转换模型2配合ρ-Ⅲ频率分布误差最小。这一结果表明：ρ-Ⅲ型分布是短历时暴雨最佳拟合线型，配合模型2，可以作为福建省城市暴雨频率转换关系式，在实际工程设计应用。本论文的研究结果对我国城市暴雨强度公式的研究具有一定的推动作用。更多还原

【Abstract】 In recent years, with the development of our country economic and the increase of the city area, the urban drainage design put forward higher requirement. The urban hydrology being the branch of the hydrology is recognised by people. Therefore, study on the urban design storm of the urban drainage engineering become one of the important problem prepared against and reduced disaster.Studies on the urban hydrology domestic and international have already made good progress, so urban drainage design is more scientific. Urban storm and storm intensity formula have been studied since the end of 70’s or the beginning of 80’s in our country. Its content is storm sampling method and parameter optimization method of storm intensity formula and storm intensity formula (including Fuzhou City, Xiamen City of the province of fujian). These researches are important to prepare against and reduce disaster and to improve city environment quality.Storm sampling method is often annual multisampling method in urban drainage engineering. But it is difficult to get the data of annual multisampling since 80 ages the sampling method of hydrology bureau is only annual maximum value. These problems need to be solved in urban drainage engineering. At the same time, with the accumulation of the data of annual maximum value from 60’s to 90’s, it provides the condition for the application of the maximum value. The data of annual maximum value gets easilier and can reduce workload and enhance efficiency. In recent years, a lot of scholars carry on studies to annual maximum method and obtain many results. Those results showed: the annual maximum method was feasible to being applied to the urban drainage engineering, but it didn’t study systematically still. As well as short duration storm has the characteristic of region, there are different characteristics in different regions.Above-mentioned, this article analyzed and calculated the data from 49 rain observation posts of Fujian’s 9 cities to look for the frequency conversion model between annual multisampling value and annual maximum value. There are four categories of storm sampling methods (annual maximum method and annual multisampling method and annual excess method and super-quantitative method). The frequency distribution model is p- III and Gumgel distribution and exponential distribution etc.. For the sake of the possibility of annual maximum method to replace annual multisampling method, we apply probability theory to built up four kinds of mathematics models. Year-frequency (1,2,3,4,5,10,15or20a) was transformed hypo-frequency according to four mathematics models. Storm design value was calculated from frequency curve of short-duration (5,10,15,20,30,45,60,90orl20mins). The error standard of assessment is criteria of sum ofabsolute value of deviations and criteria of sum of squares of deviations and criteria of NSE of the Nash- Sutcliffe. Then sum of absolute value of deviations and sum of squares of deviations and NSE were calculated in diffent short duration. Finally we can get mean value of sum of absolute value of deviations and mean value of sum of squares of deviations and mean value of NSE for 49 towns of Fujian’ 9 cities. Results showed that: (1) the error is the least when the frequency conversion model 2 fits p-III; (2) p-III is the best fitting linetype and model 2 can be the frequency conversion model of short duration storm in the province of Fujian and can be applied to engineering design. The conclusion is useful to push studies on storm intensity formula in our country.更多还原