摘要
在灰色微分动态模型的基础之上,采用季节/年际性指数对原始降水和实测径流进行预处理,并引入自记忆函数,构建灰色微分动态自记忆模型,将其应用于滦河流域径流过程的模拟和预测。结果表明:(1)采用预处理前的降水径流数据所构建的DHGM(2,2)模型和DHGM(2,2)自记忆模型在年尺度和月尺度上的径流模拟效果较差,难以反映径流的变化过程,对输入数据进行预处理后,构建的DHGM(2,2)自记忆模型模拟精度得到了很大的提高,三道河子站和滦县站年径流和月径流模拟序列的Nash-Sutcliffe系数和相关系数均达到了0.6以上;(2)模型在年尺度和月尺度的径流预测中具有一定的适用性,且结构简单、计算方便,但需要进一步考虑蒸发、土地利用和人类活动等因素,使模型更为完善。
Based on the differential hydrological grey-model, the annual/seasonal indexes were applied to data pre-proeessing of precipitation and runoff for establishing the Differential Hydrological Grey Self-memo- ry Model with the self-memory theory. The model was used to simulate and predict the runoff in different time scales. The results show that （a） DHGM （2, 2） and Self-memory DHGM （2, 2） which constructed by initial data cannot described the runoff processes in annual and monthly scales. With the pre-processing data of precipitation and runoff, the Self-memory DHGM （2, 2） performed better both in Shandaohezi and Luanxian discharges, and Nash-Sutcliffe and correlation coefficients reached more than 0.6; （b） the model was applicable to predicting the runoff especially on an annual time scale. The structure model was simple and easy to operate, nevertheless, the factors of evaporation, land-use and human activity should be taken into consideration so that the model can be more perfect,
出处
《水利学报》
EI
CSCD
北大核心
2013年第7期791-799,共9页
Journal of Hydraulic Engineering
基金
科技部基础性工作专项项目(2009IM020100、2011IM011000)
国家自然科学基金项目(51021066、51009148)
“十二五”国家科技支撑计划项目(2012BAC19B03)
中国水利水电科学研究院科研专项(资集1206)
