本文将全局参数自动优化算法—复形进化算法,引入WEP-L模型,并成功应用于汉江上游流域。通过复形进化算法参数自动优化,可找到WEP-L模型的一套较优的参数。同时,为比较不同目标函数对参数敏感性与自动率定的影响,分别对8种目标函数进行了参数敏感性分析及自动率定。结果显示,如果以水量误差和纳什效率系数为评价标准,排序后的残差平方和及其变化形式的效果较好。WEP-L模型参数敏感性分析及参数自动化率定的实现,可提高WEP-L模型参数率定的科学性和客观性,从而方便WEP-L模型的推广与应用。
Parameter optimization method of SCE-UA was successfully introduced into WEP and applied to the upstream of Han River. Through SCE-UA parameter optimization,a set of relative optimal parameter can be obtained. Moreover,in order to analyze the effect of objective function on parameter sensitivity analysis and parameter optimization,8 objective functions were used during these procedures. It was found that if water quantity error and Nash-Sutcliffe Coefficient is pursued,SSQ of sorted measured/simulated data ...