随机搜索算法在确定河流水质参数中的应用  

Chaotic Optimization for Parameter Estimation of Water Quality Model of River

作　　者：郭建青[1,2] 周宏飞[1] 李彦[1] 王洪胜[2] 

机构地区：[1]中国科学院新疆生态与地理研究所,新疆乌鲁木齐830011 [2]长安大学环境科学与工程学院,陕西西安710054

出　　处：《水文》2010年第2期25-28,共4页Journal of China Hydrology

基　　金：国家自然科学基金项目(40671037)

摘　　要：将两种随机搜索算法应用于识别瞬时投放示踪剂、一维流动的条件下的河流水质数学模型参数。分别编写了原始随机搜索和受控随机搜索两种算法的运算程序,采用不同待估水质参数的初始输入值进行了数值实验。结果表明:①两种搜索方式均可成功地应用于河流水质数学模型的参数识别问题;②待估参数的初始值输入范围对算法的收敛性几乎没有影响;③控制搜索方向对随机搜索速度有一定的改进,但改进作用并不明显。与其它算法相比,随机搜索算法具有原理非常简单、易于编程运算、搜索速度和搜索结果与参数初始输入值范围基本无关且基本不需要预先给定算法控制参数等优点。With random search algorithm,the function optimization problem from analyzing the water quality test data of river stream to estimate such water quality parameters as longitudinal dispersion coefficient,average stream velocity and the other parameter should be solved.Two ways of random search were presented and programmed in this paper.One is the primary random search,the other is the controlled random search,with different ranges of initial guessed value of water quality parameters to be input at the beginning of computation and two search ways,numerical experiment was conducted.The results show that 1） two search ways may be successfully applied to solve the function optimization problem to determine water quality parameters,2） there is little influence on convergence speed by the range of input guessed values of water quality parameters,3） controlling search direction may bring a little influence on convergence speed of random algorithm.Comparing to the other intelligence optimization algorithm,the random search algorithm is of such advantages as without key parameters of algorithm which need to be given in advance of computing,being easy to be understood and programmed to conduct computing and the precision and the results of search computing being almost no affected by the range of input guessed values of water quality parameters.

关 键 词：河流水质模型 随机搜索 参数计算 算法收敛性 

分 类 号：P342[天文地球—水文科学]