水资源与水工程学报2024，Vol.35Issue(3)：51-58,66,9.DOI:10.11705/j.issn.1672-643X.2024.03.06
南水北调受水区工业需水量的多方法组合预测
Multi-method combination prediction of industrial water demand in the receiving area of the South to North Water Diversion Project
摘要
Abstract
Based on trend analysis,multiple linear regression and BP neural network,a combined model was established to predict the industrial water demand of 2030 in the receiving area of the East Route of the South-to-North Water Diversion Project in Jiangsu Province by the relative error-inverse distance weight method with 2020 as the current year.The results show that the deviation between the predicted values of industrial water demand based on trend analysis,multiple linear regression and BP neural net-work is less than 10%,and the average error between each method and the real value is less than 10%.The coefficient of determination(R2)of industrial water demand obtained by the combined prediction model is 0.02-0.09 higher than that of any individual prediction model.The predicted value of the total industrial water demand of Jiangsu section in 2030 is 14.68×108 m3,an increase of 70.3%compared with the current year(2020).The results of this study can not only provide reliable prediction data for the East Route of the South-to-North Water Diversion Project in Jiangsu Province,but also provide a refer-ence method for industrial water demand prediction in other receiving areas of the project.关键词
工业需水预测/趋势法/回归分析法/BP神经网络/南水北调江苏段受水区Key words
industrial water demand prediction/trend analysis/regression analysis/BP neural network/Jiangsu section of the South-to-North Water Diversion Project分类
建筑与水利引用本文复制引用
毛青,解阳阳,刘赛艳,席海潮,高峥..南水北调受水区工业需水量的多方法组合预测[J].水资源与水工程学报,2024,35(3):51-58,66,9.基金项目
国家自然科学基金项目(52009116) （52009116）
江苏省高效节能大型轴流泵站工程研究中心开放课题(ECHEAP013) （ECHEAP013）
国家博士后科学基金项目(2018M642338) （2018M642338）
江苏省自然科学基金项目(BK20200958、BK20200959) （BK20200958、BK20200959）
