• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>2007 summer computer simulation conference (SCSC'07) >Using Artificial Neural Networks (ANN) for Real Time Flood Forecasting, the Omo River Case in Southern Ethiopia
【24h】

Using Artificial Neural Networks (ANN) for Real Time Flood Forecasting, the Omo River Case in Southern Ethiopia

机译：使用人工神经网络（ANN）实时洪水预报，埃塞俄比亚南部的奥莫河案例

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This study presents the application ofrnartificial neural network (ANN) methodology forrnreal time flood forecasting in Omo River,rnsouthern Ethiopia. Back propagation algorithmsrnhave been used for 1 to 6 hour runoff predictionsrnwith various combinations of flood events forrntraining the ANN models. The performance ofrneach model structure has been evaluated usingrncommon performance criteria, i.e., root meanrnsquare error (RMSE), coefficient of correlation (r),rnand coefficient of determination (R~2). The criteriarnselected to avoid over training was therngeneralization of ANN through cross validation.rnFairly accurate hourly runoff predictions havernbeen obtained using the data of six flood eventsrnsuggesting that the ANN models are particularlyrnsuited for flood forecasting purposes. The twornimportant parameters, when predicting a floodrnhydrograph, are the magnitude and the time tornpeak discharge. It has been found that the ANNrnflood forecasting models have been able to predictrnthis information with great accuracy. However,rnthe forecasting efficiency decreases withrnincreasing time.
机译：这项研究提出了人工神经网络（ANN）方法在埃塞俄比亚南部奥莫河实时洪水预报中的应用。反向传播算法已被用于1到6个小时的径流预测，洪水事件的各种组合对ANN模型进行了训练。使用通用性能标准（即均方根误差（RMSE），相关系数（r），确定系数（R〜2））对每种模型结构的性能进行了评估。为避免过度训练而选择的标准是通过交叉验证对ANN进行概括。使用六个洪水事件的数据已获得了相当准确的小时径流量预测，这表明ANN模型特别适合于洪水预报。预测洪水水文时的两个重要参数是峰值排放的大小和时间。已经发现，ANNrnflood预测模型已经能够非常准确地预测此信息。但是，随着时间的增加，预测效率下降。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号