基于支持向量机的冻融土壤入渗参数预测  

Forcast for Freezing-Melting Soil Infiltration Parameter Based on Support Vector Machine

作　　者：李昊哲[1] 樊贵盛[1] LI Haozhe FAN Guisheng(College of Hydroscience and Engineering, Taiyuan University of Technology, Taiyuan 030024, Chin)

机构地区：[1]太原理工大学水利科学与工程学院,山西太原030024

出　　处：《人民黄河》2017年第9期141-144,148,共5页Yellow River

基　　金：国家自然科学基金资助项目(40671081)

摘　　要：以提高冻融条件下农田土壤灌溉水利用率为目的,基于汾河灌区冻融期内土壤入渗试验,运用支持向量机理论,建立了以冻融条件下土壤含水率、质地、容重、地温和灌溉水温为输入变量,Kostiakov入渗模型参数k、α为输出变量的支持向量机回归预测模型。预测模型误差分析表明:入渗系数k与入渗指数α的训练样本相对误差与检验样本相对误差都控制在10%以内,将k与α代入Kostiakov入渗模型得到的90 min累计入渗量I_(90)的相对误差平均值为5.81%。说明冻融土壤条件下,基于支持向量机理论的Kostiakov入渗模型参数的预测是可行的。On the basis of the soil infiltration test during freezing inFenhe irrigation area and for improving utilization of irrigation water, SVM regression model, soil moisture content, texture, bulk density and irrigation water temperature under freezing-melting conditions were taken as input variable and infiltration parameters k and a taken as output variable, was presented by Support Vector Machines. The analysis of pre- diction error shows that the relative errors of the training samples and testing samples of both k and a are within 10%. The average relative er- ror of 190（ 90 minutes cumulative infiltration amount） of Kostiakov infiltration model is 5.81%. It shows that the algorithm is feasible for pre- dicting the parameters of Kostiakov infiltration model based on Support Vector Machines theory.

关 键 词：支持向量机 冻融土壤 Kostiakov入渗模型 土壤理化参数 

分 类 号：S152.7[农业科学—土壤学] TV93[农业科学—农业基础科学]