北京市生活用水年预报模型  

Prediction model for annual domestic water consumption in Beijing City

作　　者：顾月红[1] 葛朝霞[1] 薛梅[1] 宋颖玲[1] 

机构地区：[1]河海大学水文水资源与水利工程科学国家重点实验室

出　　处：《河海大学学报（自然科学版）》2008年第1期19-22,共4页Journal of Hohai University(Natural Sciences)

基　　金：国家自然科学基金(40233037;90211011)

摘　　要：采用偏最小二乘回归方法建立了用水量与影响因子之间的回归方程,利用AR模型对各影响因子进行外延预测,将预测的因子值分别代入3个回归模型,通过比较分析挑选出了预报精度最高的回归模型,并利用此回归模型对2003～2010年的北京市年生活用水量进行了预测.结果表明：北京市年生活用水量和前3 a的各项影响因子所组成的回归方程,能够很好地预报出预留年份的年生活用水量;2003～2010年北京市年生活用水量呈明显增长趋势.The regression equation for the relationship between water consumption and influencing factors was developed by use of the partial least square regression method. The AR model was used for prediction of each influencing factor by extrapolation, and the predicted values were introduced to 3 regression models. Then, the regression model with the highest precision of prediction was selected by comparative analysis, and it was applied to prediction of annual domestic water consumption of Beijing City from 2003 to 2010. The result shows that the regression equation based on annual water consumptions and the influencing factors in the past three years can effectively predict the domestic water consumption of Beijing City, which demonstrates an obviously increase trend from 2003 to 2010.

关 键 词：北京市 年生活用水量 影响因子 回归模型 AR模型 

分 类 号：TU991.31[建筑科学—市政工程]