煤炭工程 ›› 2017, Vol. 49 ›› Issue (8): 165-168.doi: 10.11799/ce201708047

• 工程管理 • 上一篇    下一篇

基于支持向量回归机的河北省能源消费碳排放预测

薛黎明,张心智,刘保康,胡雅各   

  1. 1. 中国矿业大学(北京)资源学院
    2. 中国矿业大学(北京)资源与安全工程学院
  • 收稿日期:2016-10-08 修回日期:2016-11-20 出版日期:2017-08-25 发布日期:2017-09-25
  • 通讯作者: 张心智 E-mail:846042148@qq.com

The SVR-based projection of Hebei’s carbon emissions resulting from energy consumption

  • Received:2016-10-08 Revised:2016-11-20 Online:2017-08-25 Published:2017-09-25

摘要: 分析了支持向量回归机在碳排放预测中的优势,并以此理论为基础构建了碳排放预测模型。采用河北省1990—2015年的碳排放量及其影响因素数据对模型进行训练和测试,得到具有良好推广能力的碳排放预测模型,并对河北省2016—2015年的碳排放量进行预测。结果表明:从增长幅度来看,1990—2015年河北省能源消费碳排放量整体呈增长趋势,增量为22 487.62万t,预测区间2016—2020年的增长量为3 055.63万t,年均增长率保持在3%左右;从增长速度来看,通过分析六项碳排放影响因素可知,人口数量、城镇化率、人均GDP以及单位GDP能耗对碳排放量增速的贡献率较小,第二产业比重和煤炭消费比对碳排放量增速的贡献率较大。最后针对得出的结论,为河北省未来几年碳减排工作提出建议。

关键词: 支持向量回归机, 碳排放, 预测模型, 河北省

Abstract: This paper analyzes the advantage of support vector regression (SVR) in the projection of carbon emission and establishes the model of projection of carbon emission based on SVR. The model with good learning and generalization ability is established through using the data of Hebei’s carbon emissions and influence factors from the year 1990 to 2013 as samples to train and test and then predicts Hebei’s carbon emissions from the year 2014 to 2020. The results show that Hebei’s carbon emissions have an increasing tendency from the year 1990 to 2020 as a whole, increasing stably and less during the period of 1990-2000 and 2010-2020 but faster from 2000 to 2010; there would be also adding up to 4441.76 tons carbon emissions in Hebei between 2014 and 2020 and keeping a stable increase rate about 3%. Finally according to the data of projection, this paper provides some suggestions which are beneficial to the Hebei’s carbon reduction.

中图分类号: