APP下载

利用Free Search算法推求作物水分响应Jensen模型参数

2018-02-01朱伟峰王斌姜宁

南水北调与水利科技 2018年6期

朱伟峰 王斌 姜宁

摘要:依据最小二乘法原理推求作物水分响应模型参数时可能存在参数不合理、模型模拟精度尚可提升、试验处理少而难于估计模型参数等问题。针对这些问题,以Jensen模型参数推求为例,引入自由搜索(Free Search,FS)算法率定作物水分响应模型参数,将FS单个动物每步探查行走的位置向量作为参数的一组潜在解,利用FS算法的动物群体迁移行为推求一组最优参数,从而改善作物水分响应模型的适用性。三个实例研究结果表明:FS算法概念清晰,操作便捷,可以直接率定作物水分响应模型参数,不需转换作物水分响应模型的数学形式,能够有效避免参数出现负值或大于1等不合理现象,并提升作物水分响应模型的模拟精度;当难于采用线性回归方法时,可以考虑利用FS算法推求作物水分响应模型参数。

关键词:作物水分生产函数;作物水分响应模型;残差平方和;参数估计;自由搜索

中图分类号:S274.1文献标志码:A

开放科学(资源服务)标识码(OSID):朱伟峰

Estimation for parameters of Jensen model using free search algorithm

ZHU Weifeng.1,WANG Bin.2,JIANG Ning.3

(1.Water Conservancy Management Center of Heilongjiang Province,Harbin 150040,China;

2.College of Water Conservancy and Civil Engineering,Northeast Agricultural University,Harbin 150030,China;

3.College of Water Conservancy and Power,Heilongjiang University,Harbin 150080,China)

Abstract:Using the least square method to estimate the parameters of the model of crop response to water (MCRW) may present some problems such as unreasonable parameters,unsatisfying simulation accuracy,and inestimable parameter due to insufficient irrigation experimental treatments.To solve these problems,Free Search (FS) was used to estimate the parameters of Jensen model in this paper.The position vector of each exploratory step of the individual animal of FS was taken as a set of potential solutions of the parameter,and a set of optimal parameters was estimated based on the animal group migration of the FS algorithm.Three case studies showed that the FS algorithm can calibrate directly the parameters of Jensen model without converting the model.FS can also effectively avoid unreasonable phenomenon such as a negative parameter or a parameter greater than 1,and improve the simulation accuracy of the Jensen model.When it is difficult to adopt the linear regression method,FS algorithm can be used to estimate the parameters of MCRW.

Key words:crop water production function;model of crop response to water;residual sum of squares;parameter estimation;free search

作物產量与水分投入量(或消耗量)之间的数学关系称为作物水分生产函数(Crop Water Production Function,CWPF),而描述作物水分生产函数的数学模型称为作物水分响应模型(Model of Crop Response to Water,MCRW)[1-4]。在大量的灌溉试验基础上,国外学者提出了一系列不同形式的作物水分响应模型,这些模型通常可以分为全生育期模型和生育阶段模型两大类,后者一般还可以细分为相加模型和相乘模型,其中常见的相加模型有Blank模型、Stewart模型、Singh模型等,而Jensen模型和Minhas模型则是较经典的相乘模型[1-2,4-8]。……

登录APP查看全文