| 用神经网络方法对雷达资料进行降水类型的分类 |
| |
| 引用本文： | 王静,程明虎.用神经网络方法对雷达资料进行降水类型的分类[J].气象,2007,33(7):55-59. |
| |
| 作者姓名： | 王静 程明虎 |
| |
| 作者单位： | 中国气象科学研究院,北京,100081 |
| |
| 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2001-2003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。
|
| 关 键 词： | 神经网络 雷达资料 降水类型 |
| 收稿时间： | 2006/3/13 0:00:00 |
| 修稿时间： | 2006-03-132007-04-28 |
Precipitation Echo Classification of Radar Reflectivity with Artificial Neural Network |
| |
| Wang Jing and Cheng Minghu.Precipitation Echo Classification of Radar Reflectivity with Artificial Neural Network[J].Meteorological Monthly,2007,33(7):55-59. |
| |
| Authors: | Wang Jing and Cheng Minghu |
| |
| Institution: | Chinese Academy of Meteorological Sciences, Beijing 100081 |
| |
| Abstract: | A Back-Propagation (BP) Model of Artificial Neural Network (ANN) is used for the partitioning of radar reflectivity into convective and stratiform-cloud precipitation classifications with the CINRAD-SA data from 2001 to 2003 in Hefei. The trained ANN is applied in a precipitation process. It is proved that the single hide-layer BP model of ANN can be used to classify the different precipitation 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 training-database, the nerve cell number of the hided-layer and the choice of the learning rate. |
| |
| Keywords: | ANN CINRAD-SA radar data precipitation classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《气象》浏览原始摘要信息 |
| 点击此处可从《气象》下载免费的PDF全文 |
