%0 Journal Article
%T 深圳市供水智能组合预测分析<br>Intelligent Combined Forecasting Analysis for Shenzhen City Water Supply
%A 郭瑜
%A 刘学智
%J Journal of Water Resources Research
%P 219-227
%@ 2166-5982
%D 2021
%I Hans Publishing
%R 10.12677/JWRR.2021.102023
%X 通过对深圳市的供水量分析，建立线性回归模型、非线性二元转折模型及模糊优选BP神经网络模型，分别对深圳市的中远期供水量进行预测，并将这三个模型的预测结果联立起来，作为模糊优选BP神经网络的输入，对深圳市供水量进行再次的网络训练。计算结果表明，智能组合预测模型的预测结果优于三个模型的单独预测结果。<br />
In this paper, water supply status of Shenzhen city is first analyzed; then linear regressive model, nonlinear binary transition model and fuzzy optimization BP neural networks model are employed to forecast medium and long-term water supply quantity, respectively. Finally, the three obtained forecasting results are combined and took as input of fuzzy optimization BP neural networks to train again, the output is just intelligent combined water supply forecasting results of Shenzhen city. Apparently, the accuracy of combined forecasting results is better than these single models.
%K 城市供水量，线性回归，非线性二元转折，模糊优选BP神经网络，智能预测<br>City Water Supply
%K Linear Regressive
%K Nonlinear Binary Transition
%K Fuzzy Optimization BP Neural Networks
%K Intelligent Combined Forecasting
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=42124