摘要
城市供水量受多种因素的共同影响。以深圳市最近20多年的供水量历史数据为基础,建立了一种基于时间序列的供水量BP神经网络预测模型。该模型的平均相对误差为4.96％。根据建立的深圳市供水量BP神经网络模型的预测结果,未来深圳市近期(2005年)的年供水量将达到52 630万m3,远期(2010年)的年供水量将达到56 142万m3。
Urban water supply quantity is influenced by several factors together. According to the historical data of water supply quantity of Shenzhen City in the last twenty years, a time line based BP neural network modei is established to forecast the urban water supplying quantity of this city. The average relative error of this modei is only 4. 96 % . According to the forecasting results of the modei, urban water supply quantity of Shenzhen City will reach to 526 300 thousand cubic meters per year in 2005 and 561 420 thousand cubic meters per year in 2010.
出处
《给水排水》
CSCD
北大核心
2004年第12期105-108,共4页
Water & Wastewater Engineering
基金
广东省科技百项工程资助项目(2KB06202S)
