| 基于BP神经网络的单站总云量预报研究 |
| |
| 引用本文： | 张长卫.基于BP神经网络的单站总云量预报研究[J].河南气象,2009(1):68-71. |
| |
| 作者姓名： | 张长卫 |
| |
| 作者单位： | 解放军理工大学气象学院; |
| |
| 摘 要： | 云是能够对飞行活动产生影响、甚至危及飞行安全的气象要素之一。为了对北京单站总云量进行预报,利用人工神经网络的BP模型,针对不同时次进行关键因子的选取,尝试用多种预报因子的组合,建立了总云量预报模型。试验结果表明：所建立的模型具有较好的拟合、预报精度,并且没有出现＂过拟合＂现象,对新样本具有较好的泛化能力。
|
| 关 键 词： | BP神经网络 单站预报 总云量 |
Single Station Forecast of Total Cloud Cover Based on BP Neural Net |
| |
| Zhang Changwei.Single Station Forecast of Total Cloud Cover Based on BP Neural Net[J].Meteorology Journal of Henan,2009(1):68-71. |
| |
| Authors: | Zhang Changwei |
| |
| Institution: | Institute of Meteorology;PLA University of Science and Technology;Nanjing 211101;China |
| |
| Abstract: | Cloud is one of the meteorological elements which have great influence on the aviation.In order to forecast total cloud cover for Beijing single station,we use the nonlinear method-BP(Back Propagation) neural net.Different key factors are chosen and composed for each foretime.The final model has a good performance in fitting and prediction.There is no "overfitting" phenomenon in the result either,so the model can be used with new samples. |
| |
| Keywords: | BP neural net single station forecast total cloud cover |
| 本文献已被 CNKI 维普 等数据库收录！ |
