应用实用动态组合模型预测城市日用水量  

Application of Practical Dynamic Combined Model for Predicting Urban Daily Water Consumption

作　　者：赵明[1] 袁一星[1] 

机构地区：[1]哈尔滨工业大学市政环境工程学院,黑龙江哈尔滨150090

出　　处：《中国给水排水》2007年第3期78-80,共3页China Water & Wastewater

基　　金：黑龙江省自然科学基金资助项目(ZJG0503)

摘　　要：将时间序列中的日用水量历史数据引入以温度等作变量的回归分析模型,建立了日用水量非线性回归组合预测模型,同时为进一步提高预测精度,用4阶自回归模型对回归残差序列进行时间序列分析,建立了日用水量预测实用动态组合模型。以华北某市日用水量的实测数据对其进行检验,结果表明该模型具有较高的预测精度。A nonlinear combined regression model for predicting daily water consumption was developed by introducing history data of daily water consumption in time series into regression analysis model using temperature and other parameters as variables. Meanwhile, to increase the prediction accuracy, the residual series were analyzed using AR （4） model, and a practical dynamic combined model for predicting daily water consumption was built. The dynamic combined model was verified using the measured daily water consumption in a city of North China region. The results show that the model has a high prediction accuracy.

关 键 词：日用水量 组合模型 预测 残差序列分析 

分 类 号：TU991.31[建筑科学—市政工程]