摘要
本文依据等流时线概念,通过将某时刻流域出口断面流量组成表达式与该时刻流域蓄水量表达式联解,得到形状类似多输入单输出的响应函数模型(简称为SMG模型),它的输入组成具有成因概念,在一定条件下,它可简化为线性变换函数模型和总线性响应函数模型。该模型的参数分析表明,其响应函数值有负数是合理的,这从成因概念上说明了响应函数不同于单位线。将SMG模型应用于龙羊峡水库汛期(5~10月)旬入库流量实际作业预报表明,其预报精度优于人工神经网络,更明显优于自回归模型和多元回归模型。
Based on the concept of channel travel isochrone, a hydrological system model in the form of response function reflecting the formation of runoff due to multiple inputs and single output is proposed. The model is deduced by jointly solving the expression of discharge at the exit of the river basin and the expression of water volume stored in the river basin at the same moment.Under the specific condition the model can be simplified as a linear transformation function model.It is applied to forecast the ten-day input discharge of the reservoir in Longyangxia Project,located at the upper reaches of the Yellow River, during flood season. The comparison of the forecasting shows that the accuracy of this model is higher than the model based on artificial neural network and other models.
出处
《水利学报》
EI
CSCD
北大核心
2004年第7期52-56,共5页
Journal of Hydraulic Engineering
基金
高等学校优秀青年教师教学科研奖励计划(高校青年教师奖)资助
关键词
成因概念
系统模型
流量预报
精确度
concept of physical cause of formation
hydrological system model
discharge forecasting, forecasting accuracy