基于PCA和改进BP网络的降雨预报模型研究  

Research in rain forecasting model based on PCA and improved BP network

作　　者：刘乐[1] 王洪国[1] 王宝伟[2] 

机构地区：[1]山东师范大学管理与经济学院,济南250014 [2]山东师范大学信息科学与工程学院,济南250014

出　　处：《计算机工程与应用》2008年第12期234-237,共4页Computer Engineering and Applications

基　　金：山东省自然科学基金(the Natural Science Foundation of Shandong Province of China under Grant No.Q2006G03)

摘　　要：在主成分分析法和改进BP网络相结合的基础上,进行降雨预报模型的研究。先由主成分分析法降低原始气象数据的维数,然后利用改进BP网络有效地学习气象样本数据中蕴含的内在规律。研究结果显示,该降雨预报模型训练效率高,预报效果好。On the base of combining Principal Component Analysis with improved BP network,this paper made a research on the rain forecasting model.First the dimensions of the raw meteorological data were decreased by PCA.Then it was by improved BP network to learn the potential rules which existd in meteorological samples effectively.The result of the research shows that,the rain forecasting model has high training efficiency and good forecasting effect.

关 键 词：主成分分析 BP网络 降雨预报 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]