中文摘要：

针对丹江口流域已有水文资料及流域气象特点,将该流域划分为12个单元区,建立降雨径流经验相关模型,应用于丹江口水库的短期洪水预报系统,并采用基于可变遗忘因子递推最小二乘算法的校正技术对预报流量过程进行实时校正。计算结果表明,基于降雨径流经验相关模型预报的洪水过程与实测过程基本一致,洪峰数值及峰现时间指标合格率较高。实践证明,模型具有较高的预报精度,可为防汛部门提供较为准确的降雨径流预报方法。

英文摘要：

On the basis of the existing hydrological data and meteorological characteristics of Danjiangkou watershed, the whole watershed is divided into 12 unit zones and the rainfall-runoff correlation model , applied to the short-term flood forecast system in Danjiangkou reservoir, is established. In addition, the correction technique based on the variable forgetting factor recursive least square algorithm is adopted to correct the flow process of the forecast in real time. The calculation results show that the predicted flood process based on the rainfall-runoff correlation model is basically the same with the actual process. Meanwhile, the pass rate of flood peak value and flood peak occurrence time is preferably high. It is proved that the forecast accuracy of this model is relatively high, demonstrating that the model can provide a more precise rainfall-runoff forecast for flood control departments.