新安江模型与衰减记忆最小二乘法耦合洪水模型的应用

Application of a Flood Model by Coupling Xin'anjiang Model with Attenuation Memory Recursive Least Squares Method

利用新安江模型进行流域产流计算,采用扩散波方程推得的汇流状态方程描述流域汇流过程,由衰减记忆递推最小二乘法实现参数估计实时化,从而使新安江模型与衰减记忆递推最小二乘法耦合,实现洪水实时预报。利用耦合模型对矶头水库76场入库洪水进行模拟,模拟结果表明,耦合模型适合中小水库入库洪水实时预报,预报精度比独立使用新安江模型得到了大大的提高。

Advantages of both Xin＇anjiang model in calculating the quantity of net precipitation, and conflux state equation which was deduced from the diffused wave equation in describing the confluence process of river flow, were taken into account in this paper. The attenuation memory recursive least squares method was used for real-time recognition of model parameters. A real-time flood forecasting model was thereby developed by coupling Xin＇anjiang Model with attenuation memory recursive least squares method. The forecasted processes of 76 floods at Jitou reservoir watershed match very well with the measured ones, showing that the coupled model in this paper is fit for real-time flood forecasting in middle or below volume reservoirs. The results also indicated that the accuracy of the coupled model in this paper is much higher than Xin＇ anjiang Model in watershed flood forecasting.