灰色模型在城市中长期用水量预测中的应用

为进行科学合理的供水系统规划,给出一种基于记录时间较短、历史数据较少的用水量序列的GM(1,1)预测方法.该预测方法把原始用水量序列累加处理生成新序列后,用指数关系式拟合,通过构造参数矩阵,确定辩识参数,建立灰色模型的微分方程;求解灰色模型的时间响应函数,生成累减矩阵,进行累减运算即得用水量序列的预测值.MAPE精度分析结果表明GM(1,1)用水量预测方法精度较高.该预测方法应用于D市的中长期用水量预测,为D市供水规划提供有效依据.

Application of GM(1,1) in long-term urban water demand forecast

To carry out the water supply system planning scientificly,the GM(1,1) forecasting method was presented based on water demand series of short record time and less historical data.After creating new series by accumulated generating the original water demand series,this forecasting method built a differential equation of gray model by fitting with exponential formula,constructing parameter matrix and ascertaining parameter identification.The forecasting value of water demand series was obtained by solving the time response function of gray model and creating inverse accumulated generating matrix for inverse accumulated generating operation.Precision analysis by MAPE method indicates that the precision of GM(1,1) is good enough to carry out a water system planning.Application of this high-precision method to the long-term water demand forecast of D city shows that it is an effective warranty for the city’s water system planning.