| 基于Lasso-GRNN神经网络模型的京津冀碳达峰情景预测 |
| 摘要点击 1726 全文点击 196 投稿时间：2023-09-01 修订日期：2024-04-26 |
| 查看HTML全文
查看全文 查看/发表评论 下载PDF阅读器 |
| 中文关键词 京津冀 碳达峰 脱钩模型 Lasso-GRNN模型 低碳发展 |
| 英文关键词 Beijing-Tianjin-Hebei carbon peak decoupling model Lasso-GRNN model low-carbon development |
|
| 中文摘要 |
| 京津冀地区作为中国能源消耗、碳排放的集聚区,降低碳排放量,实现碳达峰是该区域当下的首要目标. 以1995~2021年京津冀区域的碳排放数据以及影响碳排放的影响因素数据为研究样本,首先计算京津冀3个区域碳排放与经济增长的脱钩值,划分脱钩状态. 其次考虑到影响碳排放因素的复杂性,通过Lasso变量选择方法确定影响京津冀每个地区碳排放的关键因素,将筛选出的各个关键因素值作为GRNN与BP神经网络的输入,网络输出为对应地方的碳排放值,分析比较建立每个区域的Lasso-GRNN与Lasso-BP碳排放模型,综合各个方面分析比较Lasso-GRNN预测结果均优于Lasso-BP模型,因此选择Lasso-GRNN模型进一步设定基准情景、因素调控情景与综合调控情景进行情景分析. 结果表明：①北京市与天津市的经济增长与碳排放基本实现强脱钩,河北省整体处于弱脱钩状态,整体经济发展状态不够理想,需要进行调整优化. ②在每种情景设置下,北京市均已在2010年实现碳达峰,峰值为13 843.98万t；天津市在2013年实现碳达峰,峰值为21 115.48万t；河北省在综合因素调控情景下,将在2029年实现碳达峰,峰值为92 402.86万t. 根据研究结果对京津冀的经济发展提出合理化建议,优化产业结构,差异化发展低碳路径,使得京津冀进一步加强协作,推动低碳合作体制机制创新. |
| 英文摘要 |
| The Beijing-Tianjin-Hebei Region is a cluster of energy consumption and carbon emissions in China. Reducing carbon emissions and achieving a carbon peak are the primary goals of the region. Considering the carbon emission data of the Beijing-Tianjin-Hebei Region from 1995 to 2021, and the data on influencing factors on carbon emissions as research samples, the decoupling value of carbon emissions and economic growth in the three regions of the Beijing-Tianjin-Hebei Region was first calculated, and the decoupling state was divided. Secondly, considering the complexity of factors affecting carbon emissions, the Lasso variable selection method was used to determine the key factors affecting carbon emissions in each region of the Beijing-Tianjin-Hebei Region. The selected key factor values were considered the inputs of the GRNN and BP neural networks, and the network output was the carbon emission values of the corresponding places. The Lasso-GRNN and Lasso-BP carbon emission models of each region were analyzed and compared, and the Lasso-GRNN prediction results were superior to those of the Lasso-BP model after comprehensive analysis and comparison in all aspects. Therefore, the Lasso-GRNN model was selected to further set the baseline scenario, factor regulation scenario, and comprehensive regulation scenario for analysis and prediction. The results showed that： ① The economic growth and carbon emissions of Beijing and Tianjin achieved strong decoupling, whereas Hebei Province was in a weak decoupling state, and the overall economic development state was not ideal, which needs to be adjusted and optimized. ② Under each scenario setting, the carbon peak in Beijing was 138 439 800 tons in 2010, and Tianjin achieved the peak carbon value of 211.154 8 million tons in 2013. Hebei Province, under the comprehensive factor control scenario, was predicted to achieve the peak of carbon in 2029, with a peak of 9 240.286 million tons. Based on the research results, reasonable suggestions were put forward for the economic development of Beijing-Tianjin-Hebei, optimizing the industrial structure, and developing low-carbon paths in a differentiated way to further strengthen the cooperation between Beijing-Tianjin-Hebei and promote the innovation of a low-carbon cooperation system and mechanism. |
|
|
|
