基于BP神经网络的甘肃省二氧化碳排放预测及影响因素研究  

PREDICTION OF CARBON DIOXIDE EMISSION IN GANSU PROVINCE BASED ON BP NEURALNETWORK AND ITS INFLUENCING FACTORS

作　　者：潘思羽 张美玲[1] PAN Siyu;ZHANG Meiling(School of Science,Gansu Agricultural University,Lanzhou 730070,China)

机构地区：[1]甘肃农业大学理学院,兰州730070

出　　处：《环境工程》2023年第7期61-68,85,共9页Environmental Engineering

基　　金：甘肃省重点研发计划项目(21YF5WA096);科技部高端外国专家引进计划项目(G2022042009L);甘肃省自然科学基金项目(1606RJZA077,1308RJZA262)。

摘　　要：基于排放因子法估算2000—2020年甘肃省三大产业及生活能源消费直接CO_(2)排放量,描述分析其时序演变特征。建立BP神经网络模型并预测2021—2030年甘肃省CO_(2)排放量。构建甘肃省CO_(2)排放影响因素的STIRPAT拓展模型,利用多元回归分析定量探究了各因素对CO_(2)排放量的影响程度和内在作用机理,并结合随机森林进一步识别重要影响因素。结果表明:甘肃省产业及生活能源消耗直接CO_(2)排放总体呈波动增长趋势,且第二产业占比在70%以上,是主要的CO_(2)排放源;BP神经网络模型的预测误差为2×10-4,相关系数>0.99,对于预测甘肃省CO_(2)排放具有较高精度,并得出2026年的甘肃省能源消耗直接CO_(2)排放量达到最大;甘肃省CO_(2)排放的驱动因素作用差异显著,CO_(2)排放强度、经济发展、城乡消费对CO_(2)排放的正向作用较大,城镇居民人均消费支出是主要影响因素。森林覆盖率的增长抑制了CO_(2)排放,能源强度的降低有助于CO_(2)排放量减少,3大产业结构的调整与转型可引起CO_(2)排放量下降。The estimation of direct CO_(2) emissions from three major industries and domestic energy consumption in Gansu province from 2000 to 2020 by emission factor method was carried out,to describe and analyze its evolution characteristics.Then we established a BP neural network model and forecasted the CO_(2) emissions of Gansu province from 2021 to 2030.Finally,a STIRPAT expansion model of influencing factors of CO_(2) emission in Gansu province was constructed,the influence degree and internal mechanism of each factor on CO_(2) emissions were quantitatively explored by using multiple regression analysis,and the important influencing factors were further identified by combining random forests.The results showed that:the direct CO_(2) emission from industrial and domestic energy consumption in Gansu province were generally fluctuating,and the secondary industry accounted for more than 70%of the total CO_(2) emissions,which was the main source of carbon dioxide emissions.The prediction error of the BP neural network model was 2×10-4,and the correlation coefficient was greater than 0.99,which had high accuracy for predicting CO_(2) emissions in Gansu province,also it was concluded that the direct CO_(2) emissions of energy consumption in Gansu province would reach the maximum value in 2026.The driving factors of CO_(2) emissions in Gansu province were significantly different,the intensity of CO_(2) emissions,economic development and urban and rural consumption had a greater positive effect on CO_(2) emissions,and per capita consumption expenditure of urban residents was the main influencing factor.The growth of forest coverage inhibited CO_(2) emissions,and the reduction of energy intensity contributed to the reduction of CO_(2) emissions.The adjustment and transformation of the three major industrial structures could lead to a decline in CO_(2) emissions.The research provided a theoretical basis and scientific basis for Gansu province to promote low-carbon emission reduction.

关 键 词：CO_(2)排放 BP神经网络 STIRPAT模型 多元回归分析 随机森林 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] X321[自动化与计算机技术—控制科学与工程]