城市年用水量灰色预测研究

灰色预测模型要求原始数据序列满足指数规律,而实际上城市用水波动性大,无典型指数趋势变化,而一般呈代数曲线形式变化,因此本文提出了改进的灰色模型在城市年用水量预测中的应用,改进的灰色预测模型主要基于灰色预测模型一次累加的建模思路。将改进的灰色预测模型用于某城市的年用水量预测,结果表明:改进的灰色预测模型与传统的灰色预测模型相比,平均相对误差以及原点误差均较小,可用于该市的年用水量预测,为该市年用水的宏观调控与用水规划提供参考。

Research on grey forecasting model of urban water consumption

The gray forecasting model requests primary time-series date to meet the exponential regulation.In fact,the yearly urban water consumption date with large fluctuation do not have the typical exponential changing rule but generally have the algebraic curve changing pattern.So in this paper,an improved gray forecasting model was proposed based on the basic principles of grey model which use one-accumulated exponential model.This improved gray forecasting model was applied into one city's yearly water consumption forecasting,the result indicated that the average relative error as well as the zero point error were total smaller than ordinary gray forecasting model.It could be used in this city's yearly water consumption forecasting,and it could be given proposal to the city's water use planning and macro-economic regulating.