基于Elman神经网络的锦屏一级水电站年平均径流量集合预报研究

发布时间：2018-06-21 06:21

本文选题：锦屏一级水电站 + 项环流特征量　；参考：《水电能源科学》2017年10期

【摘要】：为预报锦屏一级水电站年平均径流量,根据其1960~1999年逐年年平均径流量与上年(1959~1998年)逐月74项大气环流指数的相关关系,选出相关性高且与年平均径流有物理联系的大气环流指数作为预报因子;利用多个Elman神经网络建立年平均径流量的集合预报模型,并采用锦屏一级水电站1960~1999年的逐年年平均径流量和筛选的预报因子数据对模型参数进行率定,2000~2011年的逐年年平均径流量对模型预报效果进行检验。结果表明,基于Elman神经网络的单一模型的范化能力较好,多模型的集合预报精度比单一模型的预报精度有进一步提高,可为锦屏一级水电站水资源调度提供参考。
[Abstract]:In order to forecast the annual average runoff of Jinping Grade I Hydropower Station, according to the correlation between the annual average annual runoff of Jinping Hydropower Station from 1960 to 1999 and the annual average runoff of last year (1959-1998), there are 74 items of atmospheric circulation index each month. The atmospheric circulation index with high correlation and physical relationship with the annual mean runoff is selected as the forecast factor, and the ensemble forecasting model of the annual average runoff is established by using several Elman neural networks. The annual average runoff of Jinping Grade I Hydropower Station from 1960 to 1999 and the selected forecast factor data are used to determine the model parameters. The annual average annual runoff from 2000 to 2011 is used to test the effect of the model forecast. The results show that the single model based on Elman neural network has better normalizing ability, and the accuracy of multi-model ensemble prediction is further improved than that of single model, which can be used as a reference for the water resources dispatching of Jinping I Hydropower Station.
【作者单位】：河海大学计算机与信息学院;中国水利水电科学研究院流域水循环模拟与调控国家重点实验室;
【基金】：十三五国家重点研发计划专题(2016YFC0402201) 十二五国家科技支撑计划项目(2015BAB07B03) 桂林市防洪及漓江补水水库群生态调度技术研究(GXZC2016-G3-2344-JHZJ)
【分类号】：P338

【相似文献】