支持向量机在三江平原井灌水稻需水量预测中的应用  

Application of Support Vector Machine to Forecast Rice Water Requirement by Well Irrigation in Sanjiang Plain

作　　者：赵清[1] 丁红[2] 刘东[2] 

机构地区：[1]农垦建三江分局水务局,黑龙江富锦156300 [2]东北农业大学水利与建筑学院,黑龙江哈尔滨150030

出　　处：《水资源与水工程学报》2010年第2期72-74,共3页Journal of Water Resources and Water Engineering

基　　金：国家自然科学基金(30400275);中国博士后科学基金(20080440832);东北农业大学科学研究基金资助

摘　　要：探讨了支持向量机在水稻需水量预测中的应用。在介绍支持向量机基本思想和回归算法的基础上,利用此方法对三江平原井灌水稻生育期内需水量进行预测。结果表明:该方法具有速度快,泛化能力强,并具有较高的模型拟合和预测精度,为当地制定合理的节水灌溉制度提供理论依据。This paper majorly discussed the application of the support vector machine SVM in rice water requirement forecast.On the foundation for introducing the basic theory of the machine and regressire algorithm method,the forecast for rice water requirement in whole grow season by well irrigation was carried out in Sanjiang Plain.The result indicated that SVM method was faster in computation and had a better generalization ability with higher model fitting and forecast precision.It can supply the theory basis local for authority to make out the reliable and precision for the development of reasonable water-saving irrigation schedule.

关 键 词：支持向量机 井灌水稻 需水量预测 三江平原 

分 类 号：S275.3[农业科学—农业水土工程]