线性回归分析在混合性产流区域短期预报的应用

Application of Linear Regression Analysis in Mixed Contributing Region Short-term Forecasts

吉尔河位于某反调节水库上游9km处,由于传播距离短,河道比降大,同时因为反调节水库的调节能力较差,为日不完全调节,壅水导致回水高程上升,吉尔站的流量传播速度由原来的1.5h缩短至0.5h,给下游的防洪调度带来了极大的难度。为进行吉尔站的短期流量预报工作,在吉尔站上游修建了一个基本水文站,在预报模型未建立初期,利用线性回归分析,假定产流条件相同,河道比降变化不大,同时忽略沿程损失,通过随即挑选数据,建立线性关系,计算相关参数,并应用到实际预报中,利用预报精度评定反复校验。

Jill River is located at the position 9 km on the upstream part of a regulating reservoir. Since propagation distance is short ; stream gradient is large, and the regulation ability of regulating reservoir is poor and belongs to daily incomplete adjustment, backwater increases the backwater elevation. The flow propagation velocity of Gil Station is shortened from 1. 5h to 0.5h, which brings great difficulty to the downstream flood control operation. A basic hydrological station was established on the upstream part of Jill Station in order to conduct short-term traffic flow forecasting work in Jill Station. Linear regression analysis was utilized in the initial stage when the forecasting model is not established. It is assumed that the runoff conditions are the same ; the river gradient change is little, and the loss along the way is ignored at the same time. Data is selected randomly to establish linear relationship and calculate the relevant parameters, which is applied to the actual forecast. And forecast accuracy evaluation is utilized for repeatedly checking.