LS-SVM在径流预测中的应用  

Application of Least Squares Support Vector Machines in Runoff Forecast

作　　者：李佳[1] 王黎[1] 马光文[1] 邬凯[2] 

机构地区：[1]四川大学水电学院,成都610065 [2]中国科学院武汉岩土力学研究所,武汉430071

出　　处：《中国农村水利水电》2008年第5期8-10,14,共4页China Rural Water and Hydropower

基　　金：国家自然科学基金项目(50539140);国家自然科学基金项目(50679098)

摘　　要：提出一种基于最小二乘支持向量机(LS-SVM)的径流预测方法。采用线性函数、多项式函数和径向基函数3种核函数进行机器学习,经过反复计算和对比分析,建立了精度较高的径流预测模型。预测实例表明,LS-SVM模型预测的平均相对误差比支持向量机(SVM)减少了2.4%,预测合格率为100%。LS-SVM建模速度快,适用于小样本情况并能得到全局最优解,将其用于径流预测是可行的。A runoff forecast method is recommended based on least squares support vector machines （LS-SVM）. Three kinds of kernel functions（linear function, poly function and radial basis function） are used in machine learning. With continued calculations and comparative analyses, a runoff forecast model with high precision is established. The practical forecasting results show that the mean percentage error of LS-SVM model decreases by 2.4% compared with SVM model and the eligibility rate is 100%. The speed of LS -SVM model is high, it can be used in limited samples to acquire the optimal solution. The proposed model for runoff forecast is feasible.

关 键 词：水利管理 径流预测 最小二乘支持向量机 核函数 

分 类 号：TV121[水利工程—水文学及水资源]