城市行业电力消耗与XCO_(2)的滚动关联挖掘方法  

Rolling Correlation Mining Method Between Urban Industry Power Consumption and XCO_2

作　　者：滕予非[1,4] 张涵 马云高 陈玉敏 张颉 刘洪利 詹宇 TENG Yufei;ZHANG Han;MA Yungao;CHEN Yumin;ZHANG Jie;LIU Hongli;ZHAN Yu(State Grid Sichuan Electric Power Research Institute,Chengdu 610000,China;State Grid Corporation of China,Beijing 100031,China;College of Carbon Neutrality Future Technology,Sichuan University,Chengdu 610000,China;Artificial Intelligence Key Laboratory of Sichuan Province,Sichuan University of Science&Engineering,Yibin 644000,China)

机构地区：[1]国网四川电力科学研究院,四川成都610000 [2]国家电网有限公司,北京100031 [3]四川大学碳中和未来技术学院,四川成都610000 [4]四川轻化工大学人工智能四川省重点实验室,四川宜宾644000

出　　处：《智慧电力》2025年第3期19-26,共8页Smart Power

基　　金：国家自然科学基金资助项目(22076129);国家电网有限公司科技项目(1400-202426279A-1-1-ZH);人工智能四川省重点实验室基金项目(2023RYY01)。

摘　　要：随着全球气候变化的加剧,减污降碳已成为全球应对环境危机的关键任务。针对当前各行业在减污降碳方面缺乏系统化、精细化技术指导的问题,提出一种城市行业电力消耗与XCO_(2)的滚动关联挖掘方法。通过结合堆叠融合随机森林、支持向量机(SVM)和极端梯度提升(XGBoost)构建高精度关联模型,进而引入时间滚动分析技术揭示不同行业电力消耗与对应XCO_(2)的动态关系。实验分析表明,关联模型性能表现优异,能为制定更精准的减排策略和能源管理措施提供支持。As global climate change intensifies,reducing pollution and carbon emissions has become a critical task in addressing worldwide environmental crises.To tackle the current lack of systematic and refined technical guidance for industries in reducing the pollution and carbon emissions,this paper proposes a rolling correlation mining method for urban industry power consumption and XCO_2.By combining stacking fusion random forests,support vector machines(SVM),and extreme gradient boosting(XGBoost),a high-precision correlation model is constructed.Furthermore,time-rolling analysis technology is introduced to reveal the dynamic relationship between the power consumption and corresponding XCO_(2) across different industries.Experimental analysis shows that the correlation model performs excellently,providing support for formulating more precise emission reduction strategies and energy management measures.

关 键 词：电力消耗 机器学习 时间滚动关联 XCO_(2) 减污降碳 

分 类 号：TM715[电气工程—电力系统及自动化]