中文摘要：

为解决流域水文模型人工试错率定参数呈现的经验性与不确定性问题,探寻模型参数率定的全局最优解,本文对三水源新安江模型与垂向混合产流模型采用复合形交叉进化算法（SCE-UA）优选率定模型参数。优化过程中拟定了两种模型参数优选边界,选定包含复合形个数p在内的SCE-UA算法参数,建立了基于洪峰、洪量、流量过程三要素的目标函数。经案例分析得出SCE-UA算法适用于上述两种水文模型调参应用,且垂向混合产流模型模拟结果优于三水源新安江模型。

英文摘要：

To solve the rate of uncertainty during the artificial trial of hydrological model and explore parameter estimation of global optimal solution, Shuffled Complex Evolution-University of Arizona （SCE-UA）algorithm is applied to optimize the parameters of both three-component Xin＇anjiang model and vertically-mixed runoff model in this paper. The optimization process has developed two-model parameters boundary and selected SCE-UA algorithm parameter chiefly containing the complex number of p and established the objective function based on three elements of peak ,magnanimity,flow process. By case analysis, the result shows that SCE-UA algorithm is applicable to both three-component Xin＇anjiang model and vertically-mixed runoff model parameters adjustment and the precision under SCE-UA optimization of vertically -mixed runoff model is higher than those of three-component Xin＇anjiang model.