%0 Journal Article
%T 基于长短时记忆神经网络( LSTM ) 的降雨径流模拟及预报
%A 廖卫红
%A 殷兆凯
%A 王若佳
%A 雷晓辉
%J 南水北调与水利科技
%D 2019
%X 长短时记忆神经网络( LSTM ) 具有很强的时间序列关系拟合能力, 非常适用于模拟及预报流域产汇流这一复杂的时间序列过程。基于 LSTM 针对不同预见期的径流预报分别建立了流域降雨径流模型, 以探讨 LSTM 在水文预报当中的应用。模型采用流域降雨、气象及水文数据作为输入, 不同预见期后的径流过程作为输出, 率定期为14 a, 验证期为 2 a。结果显示, 在预见期为 0~ 2 d 时 LSTM 预报精度很高, 在预见期为 3 d 时预报精度较差, 但仍优于新安江模型。隐藏层神经元数量作为神经网络复杂程度的代表既会影响模型预报精度, 也会影响模型训练速度。而输入数据长度则仅会在预见期为 0 的条件下影响模型预报效果。  The Long Short-Term Memory ( LSTM ) is suitable for rainfall-runoff modelling and forecasting since it has a strong ability in fitting time series. In this study , the LSTM was employed in predicting runoff in different foresight periods, in order to assess the capability of the LSTM in rainfall-runoff modelling and forecasting . The historical precipitation, meteorological and hydr ological data were used as input data, runoff at after different foresight periods were selected as model output. The calibration period is 14 years and the validation period is 2 years. As expected, the proposed model shows a great ability to predict runoff 0-2 days ahead. With 3 days of foresight period, the LSTM performs relatively poor but still better than the Xinanjiang model. The number o f hidden nodes has a primary impact on the prediction accuracy and training efficiency . While the leng th of input data has an impact on model performance only when the foresight period is 0 day .  “十三五”国家重点研发计划( 2017YFB0203104) ; 国家自然科学基金( 51709273) ; 广东省水利科技创新项目( 2017-06)
%K 降雨径流模拟
%K 水文预报
%K 机器学习
%K 深度学习
%K 长短时记忆
%K Rainfall-runoff modelling
%K Hydrological forecast
%K Machine learning
%K Deep learning
%K LSTM
%U http://www.nsbdqk.net/nsbdyslkj/article/abstract/20190601?st=article_issue