摘要
为了提高农业用水量模型的精度,以1997—2004年农业用水量数据为依据,建立了石羊河流域农业用水量的灰色GM(1,1)预测模型、三次指数平滑预测模型以及组合模型,并用2005—2007年的农业用水量数据进行检验。结果表明:组合模型的平均误差仅为0.27%,而单一的灰色GM(1,1)预测模型和三次指数平滑预测模型的平均误差分别为1.50%和0.54%。采用组合预测模型,对石羊河流域2008—2010年的农业用水量进行预测,结果分别为23.552亿、23.639亿、23.738亿m3。
In light of the data in the years from 1997 to 2004,grey prediction model GM（1,1） of agricultural water use was established as well as triple exponential smoothing forecasting model and combined model were established in order to improve the forecast accuracy of water-use model.And water consumption data during the period from 2005 to 2007 were used to test and validate the models above.The results indicates that the average error of combined model is only 0.27 %,and the single grey-exponential forecasting model GM（1,1） and three-exponential smoothing prediction model are 1.50% and 0.54% respectively.Agricultural water consumption between 2008 and 2010 is predicted by combined model in Shiyang River basin.Its results are 2.3552,2.3639,2.3738 billion cubic meters.
出处
《人民黄河》
CAS
北大核心
2012年第1期99-101,共3页
Yellow River
基金
水利部公益性行业科研基金资助项目(200801104)
