摘要
运用多种预测方法的组合预测,采用BP神经网络法、人均综合用水量法对台州市2020年、2030年总需水量进行了预测;采用人均用水量推算法、灰色预测法、数学模型法预测台州市2020年、2030年城乡生活需水量、工业需水量、第一产业需水量等分类需水量.结果表明,组合预测方法应用于台州市用水需求预测是可行的.组合预测综合考虑了各种因素的影响,能够提高需水量的预测精度,为水资源合理规划提供了科学决策依据.
The combined prediction methods based on back-propagation network and comprehensive water consumption per capita are introduced to estimate the total water demand of Taizhou in 2020 and 2030.The calculation of water consumption per capita and the method of grey forecast and mathematical model are adopted in the paper to forecast assoeted water demand in urban and rural,and that of industry and agriculture.It proves that the application of combined prediction method of water demand forecast in Taizhou is feasible,which can improve the accuracy of prediction with comprehensive consideration of various factors so as to provide scientific supports for the planning of water resources.
出处
《浙江水利水电专科学校学报》
2011年第2期37-41,共5页
Journal of Zhejiang Water Conservancy and Hydropower College
关键词
组合预测
台州
BP神经网络
需水预测
combined prediction method
Taizhou
BP network
water demand forecast
