基于主成分回归与灰色神经网络模型的水资源承载力需水量预测.docxVIP

DOI: 10． 11705 / j． issn． 1672 － 643X． 2014． 01． 22 基于主成分回归与灰色神经网络模型的 水资源承载力需水量预测 章恒全a ，何 薇b ( 河海大学 a． 商学院 决策与规划研究所; b． 商学院，南京 211100) 摘 要: 我国的水资源利用问题日趋严峻。作为影响水资源承载力的重要因素，社会经济活动对水资源的影响尤 为明显。通过分析影响水资源承载力的社会经济驱动要素，建立主成分回归模型，分析得出影响江苏省水资源承 载力变化的三个驱动力以及驱动力影响度，利用三个驱动力中的 6 个重要驱动因子，建立灰色神经网络预测模型， 预测出江苏省 2012 － 2013 年的年需水量。结果表明: 预测模型精度较高，最后结合江苏省发展现状提出相关的政 策性建议。 关键词: 水资源承载力; 主成分回归模型; 灰色神经网络模型; 需水量预测 中图分类号: TV213． 4 文献标识码: A 文章编号: 1672-643X( 2014) 01-0103-06 Forecast of water requirement of water resources carrying capacity based on regression of principal components and grey neural network model ZHANG Hengquana ，HE Weib ( a． Institute of Planning ＆ Decision － Making，Business School; b． Business School，Hohai University，Nanjing 211100，China) Abstract: Our country is facing serious water problems． As the important factor that affect water re- sources carrying capacity，the influence of social and economic activities on water resources is obvious． By analyzing the key element of socio-economic driver that affects water resources carrying capacity，es- tablishing the principal component regression model ，the driving factors which influence the carrying ca- pacity of water resources in Jiangsu were classified into three kinds ，the influence degree of the driving factors was gained． Combined with grey neural network theory and six important factors which is coming from the three main factors，the grey neural network model was established to predict the total water uses in Jiangsu Province from 2012 to 2013． The prediction accuracy of the model is higher． At last，the paper gave relative policy recommendations by combining with the actual situation of Jiangsu Province． Key words: carrying capacity of water resources; principal component regression model; grey neural net- work model; forecast of water requirement 水是人类维系生命和发展的至关重要的自然资 源。在人类发展的历史长河中，社会经济活动都是 围绕着水资源丰富的区域进行的。作为社会经济发 展的支撑性资源，水资源出现了一系列短缺和污染 等危机性现象。中国是世界上最大的