摘要
参考作物腾发量是估算作物蒸发蒸腾量的关键参数,其准确预测对提高作物需水预报精度具有十分重要的意义。最小二乘支持向量机(LS-SVM)是支持向量机(SVM)的一种改进算法,它基于结构风险最小化准则,可兼顾模型的经验风险和推广能力,将LS-SVM方法引用于参考作物腾发量预测中,并以辽宁省铁岭市为例,对比分析了LS-SVM模型与BP模型的预测结果。结果表明:LS-SVM模型学习速度快,具有比BP模型更高的模拟性能和预测精度。LS-SVM方法克服了BP模型训练时间长、容易陷入局部极小的缺点,是适合参考作物腾发量预测的新方法。
Reference evaportranspiration is a key parameter in estimating crop evaportranspiration,and its prediction is very important in estimating crop evapotranspiration and improving the use efficiency of agricultural water.The LS-SVM is a improvement of SVM algorithm. It is based on the minimum structure risk, can give dual attention to the experience risk of the modle and the promoted ability.LS-SVM was used in reference evaportranspiratien predicting, and comparison and analysis of prediction by LS-SVM model and Elman model was given in this study. The results showed that the LS-SVM model. Performed not only a quick study speed,but also a good predictable precision and stability.LS-SVM model overcomes the malpractice of BP model, more suitable in estimating crop evaportranspiration.
出处
《沈阳农业大学学报》
CAS
CSCD
北大核心
2009年第2期206-209,共4页
Journal of Shenyang Agricultural University
基金
水利部"948"项目科技创新项目(CT200516)
辽宁省教育厅技术攻关项目(02L385)
辽宁省优秀青年人才培养基金(2005230002)
辽宁省自然科学基金项目(20082122)
