求解水分运移参数的拉普拉斯-模式搜索方法  

The Method of Laplace-Pattern Search for Solving Soil Moisture Movement Parameters

作　　者：任长江[1] 白丹[1] 程鹏[2] 梁伟[3] 陈艳[1] 

机构地区：[1]西安理工大学水利水电学院,陕西西安710048 [2]华北水利水电学院数学与信息科学学院,河南郑州450011 [3]陕西师范大学旅游与环境学院,陕西西安710062

出　　处：《西安理工大学学报》2013年第1期57-60,共4页Journal of Xi'an University of Technology

基　　金：国家自然科学基金资助项目(50979089)

摘　　要：通过室内试验获得水分实测值,然后通过拉普拉斯变换对非线性水分运动方程进行线性化处理,获得水分的计算值,以土壤含水率实测值与其计算值标准差最小为优化目标,建立水分运移参数识别优化模型,采用自适应模式搜索方法对模型进行优化求解。以850、1500 min时刻实验数据作为参数识别初始数据,获取了土壤非饱和水分运移参数值,并以2 400 min时刻实验数据进行验证。含水率计算值与实测值的相关系数和标准差分别为0.946、0.022,相关系数较大,标准差较小,表明该方法识别水分运移参数是可行的。The real measured values of water contents are obtained via in-lab tests,and then the non-linear water movement equation is linearized via Laplace transformation so as to obtain the calculated values of water contents.With the real measured values of soil water contents and the minimum standard difference of calculated values as an optimized objective,the optimization model for water transport parameter identification is established.Adoptive pattern search method is used to carry out the optimization solution to the model.The experimental data in the 850,1 500 min are used as the initial data for parameter identification so as to obtain the unsaturated soil water transport parameter values,which are tested with the experimental data in 2 400 min.The correlation coefficient and standard difference between the calculated and real measured values of water contents are 0.946 and 0.022.Accordingly,correlation coefficient is high and standard difference is small,thus,indicating that this method is feasible for identifying water transport parameters.

关 键 词：拉普拉斯变换 水分运移参数 自适应模式搜索 

分 类 号：S152[农业科学—土壤学]