基于n阶多项式改进退水曲线的枯水期径流预测方法

Dry Season Runoff Forecast by Improved Recession Curve Model Based on n-order Polynomial

针对退水曲线公式中消退系数随时间逐渐递增的关系,提出用n阶多项式函数曲线拟合消退系数,构建了基于n阶多项式的改进退水曲线预测模型,通过比较各模型预测值平均偏差率选取最佳预测模型,并结合典型年法给出了枯水期径流预测的具体步骤。以澜沧江流域功果桥水电站为实例,对枯水期进行了30天的逐日径流预测,取最小平均偏差率值对应的二阶多项式改进退水曲线作为最佳模型,预测效果良好。

In view of the gradual increasing of runoff recession coefficient with time in recession curve formula, it is proposed to fit recession coefficient by n-order polynomial function, and then the improved recession curve models based on n-order polynomial are built. The optimal model is selected by comparing each average deviation rate of model, and the general forecast steps are presented combined with typical year method. Taking Gongguoqiao Hydropower Station in Lancang River basin as the example, the daily runoff forecast for 30 days in dry season is conducted. The improved recession curve model based on 2-order polynomial corresponding to minimum average deviation rate is taken as the optimal model with good forecast results.