基于主成分分析的多元线性回归用水量模型研究——以安徽省为例

用水量的准确预测对水资源规划具有重要意义。根据统计资料,选取了八项主要影响用水量变化的指标。按照主成分分析法原理,运用少数综合因子对影响用水量变化的原有八项指标进行了综合和简化,并研究其贡献率及综合效应;分析了用水量增长的特点,选用Logistic方程对综合因子进行回归分析;结合多元线性回归方法,建立了安徽省用水量模型;拟合结果表明,模型具有较高的精度;运用该模型,对安徽省2010年的用水量进行了预测;为水资源规划与管理提供参考依据。

Research on Multiple Linear Regression Water Consumption Model based on PCAM——A Case Study in Anhui Province

Veracious forecast of water consumption is very significant for the programming of water resource.According to the statistical data,eight genes,which were strongly affect on water consumption,were chosen.Bases on the principium of pain component analysis method,we used fewer colligate genes to integrate and predigest the eight genes,the different contributions and integrated response of the colligate genes were studied;After analyzing the characteristics of water consumption,Logistic equation was chosen to the regression analysis of the colligate genes;Integrate multiple linear regression method,establish the model of the water consumption of Anhui province;The result shows the high precision of the model.The water consumption of Anhui province of 2010 was forecasted.References for water resource programming and management were offered.