优选组合模型在山东省工业需水预测中的应用  

Application of Optimization of Combined Model for Industrial Water Demand Prediction in Shandong Province

作　　者：常娜[1] 陆宝宏[1] 徐玲玲[2] 刘蕊蕊[1] 阮晓波[1] 许丹[1] 

机构地区：[1]河海大学水文水资源学院,江苏南京210098 [2]江苏省水利勘测设计研究院有限公司,江苏扬州225009

出　　处：《人民黄河》2013年第3期43-46,共4页Yellow River

基　　金：国家自然科学基金资助项目(50979023);水利部公益性行业科研专项(201201026;200801003)

摘　　要：工业的迅速发展以及水资源短缺,致使工业需水量预测成为焦点。以工业生产函数法、人均综合用水量法作为基本方法分别预测了山东省的工业需水量,其中工业生产函数法考虑了参数的时变性,人均综合用水量法考虑了人口、社会等因素。应用一般的组合理论,对单一模型的预测结果进行了加权组合,通过合理性分析,人均综合用水量法是劣势因子,组合结果不好。提出并采用了先组合模型后预测的新方法,并与一般的组合方法进行比较,证明先组合模型后预测的新方法精度更高一些。最后预测了山东省2015年、2020年的工业需水量分别为212 958万、198 013万m3。The rapid development of industry and the lack of water make the industrial water demand to be the focus. The industrial water demand of Shandong Province was predicted, based on industrial productions function method, and comprehensive water consumption per capita method. The time-varying parameters in industrial productions function were joined. The population and the social factors were considered in the method of comprehensive water consumption per capita. It applied the common combined theory, and forecasted the water demand under the weights. Through rational analysis, comprehensive water consumption per capita method was worse than industrial productions function method. This method was not good. Then a new method which combined the methods then forecasted the water consumption was applied, and the practice had been proved that the precision of the new method was good. At last, it forecasted the industrial water eonsumption, which was 2 129.58 million m^3 in 2015, and 1 980.13 million m^3 in 2020 respectively.

关 键 词：工业需水量 工业生产函数法 人均综合用水量法 先预测后组合 先组合模型后预测 组合模型 山东省 

分 类 号：TV211[水利工程—水文学及水资源] X524[环境科学与工程—环境工程]