摘要
参照流域暴雨与径流相关的特点,以实际的旬初土壤蓄水量、旬降雨量为输入条件,以实际的旬径流量为输出条件,建立BP神经网络旬降雨径流相关模型。实例表明,模型结构简单,可操作性强,利用该模型进行的旬径流预报具有较好的模拟精度,并为利用旬降雨预报信息实施旬径流中期预报奠定了基础。
The ten-day correlation forecast model of rainfall and runoff based on BP neural network was established by taking actual initial soil water content and ten-clay rainfall as its inputs and the actual ten-day runoff as its output, according to the watershed characteristics of rainfall and runoff correlation. The example results showed that this model is simple in structure and easy to operation, and its ten-day runoff forecast has good simulation precision. This model laid a foundation for the ten-day runoff forecast by using ten-day rainfall forecast information.
出处
《水力发电》
北大核心
2009年第8期10-12,39,共4页
Water Power
基金
国家"十一五"科技支撑项目(2006BAB14B05)
国家自然科学基金项目(50809011)
关键词
降雨径流预报模型
旬径流
旬初土壤蓄水量
旬降雨
BP神经网络
correlation model of rainfall and runoff
ten-day runoff
initial soil water content
ten-day rainfall
BP neural network