基于LSTM的人口经济变化下交通运输碳排放预测  

Carbon Emission Prediction of Transportation Industry under Demographic and Economic Changes Based on LSTM Modeling

作　　者：马飞虎 张玉玲 谢天长 王海玲 董倩 MA Feihu;ZHANG Yuling;XIE Tianchang;WANG Hailing;DONG Qian(School of Transportation Engineering,East China Jiaotong University,Nanchang Jiangxi 330013,China;Jiangxi Tohui Science and Technology Group Co.,Ltd.,Nanchang Jiangxi 330101,China)

机构地区：[1]华东交通大学交通运输工程学院,江西南昌330013 [2]江西通慧科技集团股份有限公司,江西南昌330101

出　　处：《交通节能与环保》2025年第2期65-71,共7页Transport Energy Conservation & Environmental Protection

基　　金：国家重点研发计划项目(2021YFE0105600);国家自然科学基金面上项目(51978263);江西省自然科学基金重点项目(20192ACBL20008)。

摘　　要：为研究人口经济变化下交通运输业碳排放情况,本文深入探究区域碳排放量以及人口经济、能源消费量的现状,分析区域“十二五”和“十三五”期间的碳排放总量和变化趋势。基于统计学理论,引入Person相关性,建立多元线性回归模型,确定碳排放预测模型参数取值。建立线性模型求出人口与经济对能源消费量的影响因子,并使用该模型对未来的能源消费量进行预测。根据2010—2020年的指标数据建立LSTM预测模型并进行评估,利用LSTM模型预测2021—2060年的碳排放量。研究结果表明,2010—2020年交通碳排放量总体呈上升趋势,未来能源消耗量持续上升,增长率变缓,在不进行任何人工控制下,未来碳排放量将呈现阶梯上升的趋势。In order to study the carbon emissions from the transportation industry under the demographic and economic changes,the current situation of regional carbon emissions as well as demographic and economic and energy consumption is explored in depth,and the total amount of carbon emissions and the trend of changes during the 12th and 13th Five-Year Plan periods are analyzed in the region.A multiple linear regression model is established by introducing Person correlation to determine the values of the parameters of the carbon emission prediction model.A linear model is established to find out the influence factors of population and economy on energy consumption,and the model is used to forecast future energy consumption.The LSTM prediction model is established and evaluated based on the indicator data from 2010 to 2020,and the LSTM model is used to predict the carbon emissions from 2021 to 2060.The results of the study show a general upward trend in transportation carbon emissions from 2010 to 2020,with future energy consumption continuing to rise and the growth rate becoming slower,and a stepwise upward trend in future carbon emissions without any artificial control.

关 键 词：交通运输 LSTM 碳排放预测 多元线性回归 

分 类 号：X322[环境科学与工程—环境工程]