| 一种改进的循环神经网络雷达图像外推算法 |
| |
| 引用本文： | 尹麒名,甘建红,漆慧,胡文东,张莹,黎仁国,唐旺.一种改进的循环神经网络雷达图像外推算法[J].气象科技,2021,49(1):18-24. |
| |
| 作者姓名： | 尹麒名 甘建红 漆慧 胡文东 张莹 黎仁国 唐旺 |
| |
| 作者单位： | 成都信息工程大学，成都 610200;西华师范大学，南充 637002 |
| |
| 基金项目： | 四川省基础应用研究重点项目（2018JY0056）、四川省应用基础研究（2019YJ0361）、气象信息共享与数据挖掘四川省高校重点实验室开放课题（QGX16009）、气象信息共享与数据挖掘四川省高校重点实验室课题(QGX18004)、四川省教育厅一般项目(16ZB0222)、四川省应用基础研究（2020YJ0425）、成都市重点研发支撑计划（2019 YF05 00219 SN）资助 |
| |
| 摘 要： | 临近强降水预报目的是预测未来两小时内局地降水强度的分布,准确的外推雷达图像可以为临近强降水预报提供准确的时空参考数据。近两年循环神经网络模型应用于天气雷达回波图象外推得到了较好的结果。本文基于分析现有ConvLSTM和TrajGRU模型的基础上,从输入雷达数据层数和修改模型损失函数两个方面对循环神经网络外推模型进行改进,并对业务上的雷达图象序列和竞赛雷达图象序列进行试验。试验结果表明,改进的外推模型能更好地捕捉时空相关性,具有更精确的外推效果。
|
| 关 键 词： | 临近预报 雷达回波外推 深度学习 门控循环单元 长短记忆模型 |
| 收稿时间： | 2020/1/17 0:00:00 |
| 修稿时间： | 2020/8/25 0:00:00 |
An Improved Algorithm of Radar Image Extrapolation Based on Recurrent Neural Network |
| |
| YIN Qiming,GAN Jianhong,QI Hui,HU Wendong,ZHANG Ying,LI Renguo,TANG Wang.An Improved Algorithm of Radar Image Extrapolation Based on Recurrent Neural Network[J].Meteorological Science and Technology,2021,49(1):18-24. |
| |
| Authors: | YIN Qiming GAN Jianhong QI Hui HU Wendong ZHANG Ying LI Renguo TANG Wang |
| |
| Institution: | (Chengdu University of Information Engineering,Chengdu 610200;China West Normal University,Nanchong 637002) |
| |
| Abstract: | The purpose of precipitation nowcasting is to predict the distribution of local precipitation intensity within the coming two hours, and accurate extrapolation radar images can provide accurate space time data reference for nowcasting. The application of Recurrent Neural Network (RNN) to meteorological radar image extrapolation has brought better results in recent two years. Based on the analysis of convLSTM and TrajLSTM, the extrapolation model is improved from two aspects: the number of layers of radar data and the loss function, and the experiment is carried out with Chengdu and Yueyang radar data and an open competition data set. The experimental results show that the improved model can better capture the space time correlation and keep more image details. |
| |
| Keywords: | nowcasting radar image exploration deep learning GRU LSTM |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《气象科技》浏览原始摘要信息 |
| 点击此处可从《气象科技》下载免费的PDF全文 |
|
