| 基于前传式网络逼近的太平洋副热带高压活动的诊断预测 |
| |
| 引用本文： | 张韧.基于前传式网络逼近的太平洋副热带高压活动的诊断预测[J].大气科学,2001,25(5):650-660. |
| |
| 作者姓名： | 张韧 |
| |
| 作者单位： | 解放军理工大学气象学院海洋气象教研室南京 211101 |
| |
| 基金项目： | 国家自然科学基金资助项目49975012及空军司令部气象局(司气字第45号)共同资助 |
| |
| 摘 要： | 基于前传式神经网络BP算法(BackpropagationNeuralNetwork)和回归模型,探讨了西太平洋副高面积指数同赤道东太平洋海温及赤道纬向风之间非线性分类和映射逼近的建模方法和效果比较.结果表明,前传式网络,特别是回归网络预报模型具有较好拟合精度和预报效果及比较实用的预报时效.
|
| 关 键 词： | 前传式网络 回归模型 副热带高压 |
A Diagnosis/ Prediction of the Movement of the Western Pacific Subtropical High Based on the Neural Network Approximation with Forth-Propagation Style |
| |
| Zhang Ren.A Diagnosis/ Prediction of the Movement of the Western Pacific Subtropical High Based on the Neural Network Approximation with Forth-Propagation Style[J].Chinese Journal of Atmospheric Sciences,2001,25(5):650-660. |
| |
| Authors: | Zhang Ren |
| |
| Abstract: | Based on the BP arithmetic and regression model of the neural network with forth--propa- gation style, discussed are a modeling method and effect comparison of non-linear classification and mapping approximation between the western Pacific subtropical high area index and the equatorlal eastern Pacific SST / the equatorial zonal winds. The results show that the neural network models with forth--propagation style (especially, the regression network model) are able to forecast with better fitting--precision and timely forecasting effectiveness. |
| |
| Keywords: | neural network with forth-propagation style regression model subtropical high |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《大气科学》浏览原始摘要信息 |
| 点击此处可从《大气科学》下载免费的PDF全文 |
