摘要
针对传统的碳排放测算模型因缺乏对大气污染因素的考虑,导致测算结果存在偏差的问题,提出大气污染约束下建筑物全生命周期碳排放测算模型。基于归因模型与注意力机制对采集到的用于后续建筑物碳排放测算的数据展开缺失值填充;通过地理加权回归(GWR)计算大气污染约束下的碳排放因子值,使得后续碳排放测算结果更加贴近实际情况;依据填充后的建筑物数据设计基于全生命周期的建筑物碳排放量测算模型,将碳排放因子代入模型完成建筑物碳排放量测算。结果表明,所提方法的建筑物全生命周期碳排放测算准确度更高,更适用于实际应用。
Aiming at the problem of deviation in the calculation results caused by the lack of consideration of atmospheric pollution factors in traditional carbon emission calculation models,a building lifecycle carbon emission calculation model is proposed under atmospheric pollution constraints.The study used attribution models and attention mechanisms to fill in missing values in the data collected for subsequent building carbon emission calculations.By using Geographically Weighted Regression(GWR)to calculate the carbon emission factor values under atmospheric pollution constraints,the subsequent carbon emission measurement results can be more closely aligned with the actual situation.Based on the filled building data,the study designed a building carbon emission calculation model based on the entire life cycle,and input carbon emission factors into the model to complete the calculation of building carbon emissions.The results indicate that the proposed method has higher accuracy in measuring the carbon emissions of buildings throughout their entire lifecycle and is more suitable for practical applications.
作者
吝含伟
张玉婧
Lin Hanwei;Zhang Yujing(Guangzhou Urban Planning Design Survey Research Institute Co.Ltd,Guangzhou 510060,China;Guangzhou Collaborative Innovation Center of Natural Resources Planning and Marine Technology,Guangzhou 510060,China;Guangdong Enterprise Key Laboratory for Urban Sensing,Monitoring and Early Warning,Guangzhou 510060,China)
出处
《环境科学与管理》
2025年第8期33-38,共6页
Environmental Science and Management
基金
广州市资源规划和海洋科技协同创新中心项目(2023B04J0301,2023B04J0046)
广东省重点领域研发计划资助(2020B0101130009)
广东省城市感知与监测预警企业重点实验室基金项目(2020B121202019)资助。
关键词
全生命周期
地理加权回归模型
碳排放因子
碳排放量
建筑物
full lifecycle
geographically weighted regression model
carbon emission factor
carbon emissions
buildings
