摘要
目的利用集对分析思想,构建集对分析聚类预测模型对国内未来生活用水量进行长期预测,分析未来生活用水量及增长趋势,为有关部门决策提供参考意见。方法从2000年至2014年的《中国统计年鉴》中收集年居民生活用水量时间序列数据,利用聚类的思想,将2000年至2013年的生活用水量增长率分成3类,利用集对分析思想,计算居民生活用水量与影响因子人口总数的联系度,通过预测人口总数在未来的增长率,预测未来居民生活用水量及其增长速度。结果从聚类的结果可以看出,2001-2013年生活用水量除2012年有所回落外,一直呈增长态势,年增长率除2012年外均稳定在1.02~1.04之间;由预测的结果来看,2014-2025年的生活用水量均属于第二类,即增长率范围稳定在[0.9833,1.0273],到2025年生活用水量将达到869.06亿m3;将2010-2013年实际生活用水量增长率与预测生活用水量增长率进行比较,计算其预测误差均在2%以内。结论集对分析聚类预测模型对于生活用水量的预测具有较高的精度,可以较好地预测长期生活用水量。未来生活用水量呈现缓慢增长态势,在一定增长率水平下,2025年突破850亿m3。
Objective According to the set pair analysis theory,to construct clustering set pair analy-sis prediction models to make long-term forecasts for domestic water consumption in the future,and ana-lyze domestic water consumption and its growth trends to provide reference for the decision-making of rele-vant departments.Methods The household water consumption time series data from China Statistical Yearbook of 2000 to 2014 were collected.By using the idea of clustering,the growth rate of domestic wa-ter consumption from 2000 to 2013 was divided into three categories.And then by using the set pair analy-sis,the connection degree between the water consumption and the influence factor of the population were determined.Finally,the future water consumption and its growth rate were predicted by predicting the future growth rate of the population in the future.Results The results of the clustering showed that wa-ter consumption in 2001-2013 (except in 2012),was in growing trend,and the annual growth rate was stable between 1.02-1.04 (except in 2012).The predicted results showed that living water consumption of 2014-2025 years would belong to second category,the growth rate range would be stable in the interval of [0.9833,1.0273],living water consumption would reach 86.906 billion cubic meters.The growth rate of the real living water consumption of 2010-2013 was compared with the forecast of the growth rate of domestic water consumption,the prediction errors were less than 2% by calculating.Conclusion There-fore,the set pair analysis model has high precision for forecasting the water consumption,and long term water consumption can be predicted well.Future living water consumption will be in a slow growth trend <br> and break 85 billion cubic meters in 2025 under a certain growth rate.
出处
《河北北方学院学报(自然科学版)》
2016年第3期32-36,共5页
Journal of Hebei North University:Natural Science Edition
基金
国家自科科学基金项目(11301001)
安徽财经大学教研项目(acjyyb2014104)
关键词
生活用水量预测
集对分析法
聚类预测
联系度
living water consumption forecasting
set pair analysis
clustering prediction
connection degree