【作者】 李宏伟；

【导师】 宋松柏；

【作者基本信息】 西北农林科技大学 ， 水文学及水资源， 2009， 硕士

【摘要】 水文频率分析是水文学研究中的一项重要内容,其主要目的是通过使用概率分布来揭示水文事件的数值与它们发生频率的相互关系,以便为水利工程规划,水资源优化配置等提供依据。本文查阅了近年来国内外水文频率研究的文献,总结了水文频率分析的研究现状,介绍了现有水文频率分布模型及其参数估计方法,并在此基础上对频率分布模型、常规参数计算方法、智能参数计算方法进行了分析研究,研制开发了通用频率计算软件。本文的主要研究内容如下:(1)综述了近些年来国内外水文频率研究进展,回顾了水文频率分析的现有方法,总结了水文频率研究的最新趋势,介绍了正态分布类、Г分布类、极值分布类、Wakeby分布类、Logistic分布类五大分布类共14种分布的概率分布函数及其常规参数计算方法(矩法(MOM)、极大似然法(ML)、概率权重矩法(PWM))。(2)介绍了模拟退火算法(SA)、遗传算法(GA)、粒子群算法(PSO)和蚁群算法(ACA)这四种智能优化算法的基本原理及算法流程,在此基础上建立了水文频率曲线线型优化模型,并通过编制计算机程序实现了水文频率分析中参数计算过程的自动化。(3)编制了一套通用水文频率分析系统,该系统包含了正态分布类、Г分布类、极值分布类、Wakeby分布类、Logistic分布类五大分布类共14种概率分布的分布模型,并集成了3种线型优选准则(离差平方和最小准则(OLS)、离差绝对值和最小准则(ABS)、相对离差平方和最小准则(WLS))和对应上述14种分布的常规参数计算方法(包括矩法(MOM)、极大似然法(ML)、概率权重矩法(PWM)),在此基础上,应用模拟退火算法(SA)、遗传算法(GA)、粒子群算法(PSO)和蚁群算法(ACA))进行水文频率分布参数的优化计算,该系统可以作为各种水文频率分析的一套基础性研究工具,后续相关研究工作可以以此为基础展开。(4)以陕北地区12个测站的年径流系列为基础资料,通过通用频率分析系统对每个测站的资料分别按12种不同分布,各分布采用7种参数计算方法,计算了分布参数,通过统计分析,得出了适用于陕北地区年径流系列的理论频率分布模型和最佳参数计算方法以及不同设计频率下的年径流值。更多还原

【Abstract】 Hydrologic frequency analysis is an important part of hydrology research, its main purpose is to show the relations between the value of hydrologic events and their happen frequency using the probability distribution, and supply for the hydraulic engineerings planning and the water resourses distribution.This thesis reviewed the literatures of hydrologic frequency analysis, summarized the the present progresses, introduced the distribution models and methods of estimating parameters of hydrologic frequency, and base on them, analysed the frequency distribution models, traditional methods of estimating parameter, intelligent methods of estimating parameter, then developed a software for hydrologic frequency analysis. The major results of the thesis is as follows.(1)The domestic and overseas research progress of hydrologic frequency analysis in recent years is summarized, the present methods of hydrologic frequency analysis is reviewed, research trend of hydrologic frequency is presented. The probability distribution function and its routine methods of estimating parameter are summarized, the distributions includes that normal distribution class,Гdistribution class, extreme value distribution class, Wakeby distribution class and Logistic distribution class, the routine methods contains moment method, maximum likelihood method and probability-weighted moment method.(2)The basic theory and algorithm flow of four intelligent algorithms such as Simulated Annealing Algorithm, Genetic Algorithm, Particle Swarm Optimization and Ant Colony Algorithm is introduced, the optimization models of hydrologic frequency curve are used to hydrological frequency analysis, and they have been implemented in computer to make the parameter estimating of hydrologic frequency analysis.(3)A software for general hydrologic frequency analysis is developed, it includes fourteen distribution models(normal distribution class,Гdistribution class, extreme value class, Wakeby distribution class and Logistic distribution class), integrates three curve optimization rules(OLS, ABS, WLS) , the routine methods(MOM, ML, PWM) and intelligent methods(SA, GA, PSO, ACA) of estimating parameters. This system can be used a basic tool for the hydrologic frequency analysis, some subsequent research can base on it.(4)Based on the annual runoff data of north region of Shaanxi, the parameters of different sites of different distributions have been calculated using different parameter methods, the best theory frequency distribution, best method of estimating parameters and annual runoff of different frequency have been obtained by the software.更多还原