基于套索算法和灰色模型的浙江省碳排放量分析与预测  

Carbon emission forecasting in Zhejiang Province based on LASSO algorithm and grey model

作　　者：洪竞科 杜薇 邵金 劳慧敏[2] HONG Jingke;DU Wei;SHAO Jin;LAO Huimin(School of Management Science and Real Estate,Chongqing University,Chongqing 400044,China;Institute of Science and Technology Information of Zhejiang Province,Hangzhou 310006,China)

机构地区：[1]重庆大学管理科学与房地产学院,重庆400044 [2]浙江省科技信息研究院,浙江杭州310006

出　　处：《能源环境保护》2024年第3期152-161,共10页Energy Environmental Protection

基　　金：浙江省“尖兵”“领雁”研发攻关计划资助项目(2022C03146);国际合作重点资助项目(T2261129477);教育部哲学社会科学研究重大课题攻关资助项目(21JZD029);浙江省重点软科学研究资助项目(2023C25059)。

摘　　要：在我国“碳达峰、碳中和”的绿色低碳发展背景下,对碳排放量的周期性分析与准确预测具有重要的现实意义。以浙江省碳排放量为研究对象,引入变分模态分解方法提取多尺度信息,将套索算法(Least Absolute Shrinkage and Selection Operator,LASSO)和灰色模型GM(1,N)相结合,对浙江省的碳排放量进行深入分析和预测。首先,使用变分模态分解方法对浙江省碳排放量进行数据分解,分析其历史波动的周期性。其次,利用套索算法对影响碳排放的关键因素进行有效筛选,降低数据维度,提取主要特征。最后,结合“十四五”规划与实际发展路径,假设了常态、低碳、惯性发展情景,并采用GM(1,N)模型对浙江省2020—2030年碳排放量进行预测,克服了传统预测方法在处理非线性、小样本数据时的局限性,预测结果更加稳健。结果表明浙江省的碳排放量的主导因素包括第三产业占GDP的比重、私人汽车拥有量、全省固定资产投资、电力消费总量、研发强度、技术市场成交额6个指标。预计到2030年,仅低碳发展情景下的碳排放量达到峰值,为400.28 Mt,而常态发展情景的碳排放量为474.23 Mt,惯性发展情景为568.77 Mt,且常态发展情景和惯性发展情景下的碳达峰量在2030年后仍会有所增长。因此,建议浙江省继续优化产业结构、提升能源效率、增加低碳研发投入,稳扎稳打推进“碳达峰”目标。Under the green and low-carbon development goal of achieving"carbon peaking and carbon neutrality"in China,cyclical analysis and accurate prediction of carbon emissions are of great importance.This paper investigates carbon emissions in Zhejiang Province.First,the variable mode decomposition method is used to decompose the historical data of carbon emissions in Zhejiang Province,enabling an analysis of its cyclicality fluctuations.Second,the LASSO algorithm is employed to identify the key influencing factors of carbon emissions.Finally,considering the 14th Five-Year Plan and the province′s development trajectory,three development scenarios(normal,low-carbon,and inertia)are assumed,and the GM(1,N)model is used to predict the carbon emissions in Zhejiang Province from 2020 to 2030.The analysis reveals that the dominant factors affecting carbon emissions in Zhejiang Province are the proportion of the third industry in GDP,the number of private cars,the total fixed asset investment in the province,the total electricity consumption,R&D intensity,and technology market turnover.Under the low-carbon scenario,carbon emissions are projected to peak at 400.28 Mt in 2030.In contrast,under the normal scenario,carbon emissions are estimated to reach 474.23 Mt,while the inertia development scenario predicts carbon emissions of 568.77 Mt.Furthermore,carbon emissions are expected to continue rising beyond 2030 in the normal and inertia development scenarios.In light of these findings,It is recommended that Zhejiang Province should focus on optimizing its industrial structure,improving energy efficiency,increasing investment in low-carbon research and development,and steadily advancing the goal of"carbon peaking".

关 键 词：碳排放量 套索算法 GM(1 N) 预测 

分 类 号：X32[环境科学与工程—环境工程]