摘要
对江西乐安河流域香屯站1956年-2005年年最高水位,采用均生函数方法,生成25个周期性基函数,利用SPSS软件进行逐步回归因子挑选,最终筛选出10个周期性基函数作为预测对象的影响因子,建立最优回归方程。结果表明:该模型对非极值的模拟和预测精度很高,对极值的模拟效果较其它模型有很大改善,但仍然是主要误差来源。
The method of average-growing function was performed with maximum annual water level at Xiangtun station from the year of 1956 to 2005 to create 25 periodic functions. After filtering by stepwise regression factor method employed by SPSS software, the final 10 periodic functions were selected as the impact factors to predict and establish the optimal regression equation. The results showed that the model had a high precision to predict and simulate the non-extreme value. The result of extreme values had significantly improved compared with other models, however, which were still attributed to the main source of error.
出处
《南水北调与水利科技》
CAS
CSCD
2010年第1期72-74,共3页
South-to-North Water Transfers and Water Science & Technology
关键词
均生函数
周期性基函数
SPSS
逐步回归
average-growing function
periodic functions
SPSS
stepwise regression
作者简介
黄燕（1985-），男，湖北监利人，主要从事水文水资源方面的研究工作。
