基于最小二乘支持向量机的龙羊峡径流预测  

Longyangxia Runoff Forecasting Based on LS-SVM

作　　者：陶凤玲[1] 武晟[2] 余生才[1] 肖博[2] 

机构地区：[1]青海大学水电系,青海西宁810016 [2]西安理工大学水电学院,陕西西安710048

出　　处：《水文》2008年第4期29-30,23,共3页Journal of China Hydrology

基　　金：青海科技厅重点攻关项目“龙羊峡水电厂发电损失的补偿机制研究”

摘　　要：本文简述了支持向量的回归分析、支持向量机自回归预测模型结构及预测方法,利用黄河唐乃亥站47年的径流资料,采用最小二乘支持向量机方法对径流进行了模拟预测,并与B-P神经网络方法进行了对比分析,其计算结果相对略好。This paper introduced the support vector regression analysis, expounded on support vector machine auto-regression model structure and prediction methods. The runoff data of 47 years were taken for runoff prediction by using least square method. A contrast analysis was made with the B-P neural network method, which showed that the computed result was relatively better than the B-P neural network method.

关 键 词：支持向量机 径流预测 BP神经网络 龙羊峡 

分 类 号：P338.9[天文地球—水文科学]