本文已被：浏览 873次 下载 1942次
投稿时间：2006-03-13 修订日期：2007-04-28
投稿时间：2006-03-13 修订日期：2007-04-28
中文摘要: 利用不依赖先验统计模型的多层前馈神经网络模型对合肥的新一代S波段A系列雷达20
01—2003年的降水资料进行了三种降水类型的分类，并将训练完成后的网络应用于一次降水
过程。利用单隐层的多层前馈神经网络模型，在取适当参数时，已经可以较好地对雷达资料
进行对流云降水、层状云降水和混合云降水三种降水类型的分类。同时验证了：训练集样本
的数量和顺序、隐层神经元的数目以及学习率的选择等都将影响分类的成功率。
Abstract:A Back Propagation (BP) Model of Artificial Neural Network (ANN) is used for th
e partitioning of radar reflectivity into convective and stratiform cloud preci
pitation classifications with the CINRAD SA data from 2001 to 2003 in Hefei. Th
e trained ANN is applied in a precipitation process. It is proved that the singl
e hide layer BP model of ANN can be used to classify the different precipitatio
n echoes with a high success rate. It is also validated that: the success rate
is influenced by following factors: the amount and the in put order of the tr
aining database, the nerve cell number of the hided layer and the choice of th
e learning rate.
文章编号： 中图分类号： 文献标志码：
基金项目:
| Author Name | Affiliation |
| Wang Jing | Chinese Academy of Meteorological Sciences, Beijing 100081 |
| Cheng Minghu | Chinese Academy of Meteorological Sciences, Beijing 100081 |
引用文本：
王静,程明虎,2007.用神经网络方法对雷达资料进行降水类型的分类[J].气象,33(7):55-59.
Wang Jing,Cheng Minghu,2007.Precipitation Echo Classification of Radar Reflectivity with Artificial Neural Network[J].Meteor Mon,33(7):55-59.
王静,程明虎,2007.用神经网络方法对雷达资料进行降水类型的分类[J].气象,33(7):55-59.
Wang Jing,Cheng Minghu,2007.Precipitation Echo Classification of Radar Reflectivity with Artificial Neural Network[J].Meteor Mon,33(7):55-59.