基于主成分回归分析的需水预测研究  

Research on the Water Demand Prediction based on Principal Component Regression

作　　者：张少杰 游洋[2] 

机构地区：[1]南京市长江河道管理处,江苏南京210011 [2]江苏省水利厅,江苏南京210029

出　　处：《海河水利》2016年第3期43-45,56,共4页Haihe Water Resources

摘　　要：水资源是城市发展的动力,需水量准确预测对城市可持续发展具有重要意义。需水量受多重因素影响,单一使用多重线性回归难以保证预测的准确性和科学性。根据南京市2005—2014年7个经济、社会发展相关指标,利用主成分回归分析建立模型使用原始变量对用水量进行预测。结果表明,应用主成分回归模型进行需水预测,比多重线性回归模型精度高,也较好地拟合了实际用水量。Water resources are the driving force of urban development. Accurate predictiong of water demand is very impor- tant for urban sustainable development. Water demand is influenced by multiple factors, and the accuracy and the scientific using multiple linear regression is difficult to ensure the prediction. According to 7 factors of Nanjing 2005-2014, the mod- el was created to predict the water consumption by principal component regression analysis. The results show that the princi- pal component regression model is more accurate than the multiple linear regression model, and the principal component re- gression model is more fit the actual water consumption.

关 键 词：水资源 需水预测 主成分 回归 南京 

分 类 号：TV213.4[水利工程—水文学及水资源]