基于影响因素分析的需水预测研究
2018-07-28张天一刘心
张天一 刘心
摘要:随着经济的增长和人民生活水平不断地提高,国内生活用水需求量呈现逐年增长的趋势,水资源短缺已成为人类发展的瓶颈,合理利用水资源势在必行。需水预测环境因素复杂,不同的地区影响因子各不相同,且影响因子的筛选直接决定需水量预测的结果精确与否。对此,本文提出了考虑节假日、降雨、季节、高温天气四项因素与BP神经网络、时间序列预测相结合的需水预测模型。并以校园用水为例,结合实际环境,建立模型,进行预测。结果表明,本文所提出的方法能够有效的预测校园需水量,预测精度较于单一的预测模型有明显提高。
关键词: BP神经网络1;时间序列2;预测3;影响因素4;用水5
中图分类号:TP18 文献标识码:A 文章编号:1009-3044(2018)11-0283-05
Water Demand Forecasting Based on Analysis of Influencing Factors
ZHANG Tian-yi, LIU Xin
(Hebei Engineering University, Hebei Handan, 056038, China)
Abstract:With the economic growth and improvement of people's living standard, the demand for domestic water has been increasing year by year. The shortage of water resources has seriously endangered the normal life of human beings and the rational use of water resources is imperative. There are many factors involved in water demand forecastin. The influence factors in different regions are different and vary, and the choice of influence factors directly determines the result of water demand forecasting. A water demand forecasting model combining four factors of holidays, rainfall, season and high temperature with BP neural network and time series forecast is proposed. Take the campus water as an example, through forecasting the main influencing factors and constructing the water demand forecasting model with the influence of factors.
Key words:BP neural network1; time series2; water demand prediction3; influencing factors4; water5
1 引言
水資源日益短缺的今天,人类社会生存和发展无可取代的自然资源及生态环境系统的组成就是宝贵的水资源,合理利用水资源已是势在必行。需水预测极为复杂,主要是因为涉及的影响因素众多,如人口、社会经济发展水平、节假日、高温天气、降水量等。需水量预测的准确程度直接影响供水系统投资、管网布局和运行的合理性[1-3]。不论选取什么预测方法,首要考虑的就是合理分析并筛选各因子作为输入模型,保证预测精准度。因此,要想合理使用宝贵的水资源、进行有效的供水就必须对水资源进行科学严瑾的预测。
当前需水预测的方法有很多,较早的需水预测方法有定额法,它的准确度较低,由实验经验决定[4]。……
