摘要
在减污降碳协同推进的新目标形势下,实现减污降碳协同增效是实现高质量发展的关键。多目标优化法在协同减污降碳领域中具有重要的应用价值。从当前工艺中多目标优化的4个主要步骤:决策变量与目标函数、机器学习方法构建、多目标寻优、最优解选取进行综述,得出决策变量选取不仅要考虑强耦合关系的可控变量,还应注重非可控变量的影响。在机器学习方法构建和求解时,应根据工艺特征、样本数量、目标维度等选择适用的算法并对其局限性进行验证和改进。未来应加强对污染物和碳排放的同源性和异质性识别,完善数据质量和机器学习方法精度,开展Pareto支配关系修正和Pareto解集相关性研究,提升寻优速度。
With the increasing severity of environmental pollution and carbon emissions,the reduction of pollution and carbon emissions in process technology has become an important research direction.Based on multi-objective optimization method has become an important way to solve the problem of pollution reduction and carbon reduction in the process.Summarize the four main steps of multi-objective optimization in the current process:decision variable selection,model construction,multi-objective optimization,and optimal solution selection.Concluded that the selection of decision variables should not only consider controllable variables with strong coupling relationships,but also pay attention to the impact of uncontrollable variables.In model construction and solution,appropriate algorithms should be selected based on process characteristics,sample size,target dimensions,and their limitations should be verified and improved.In the future,it is necessary to strengthen the identification of the homology and heterogeneity of pollutants and carbon emissions,improve data quality and model accuracy,conduct research on the modification of Pareto dominance relations and the correlation of Pareto solution sets,and enhance optimization speed.
作者
何咏
李金玉
HE Yong;LI Jing-yu(Beijing Municipal Research Institute of Eco-Environmental Protection,Bejing 100037,China;National Engineering Techniques Research Center of Urban Environmental Pollution Control,Beijing 100037,China)
出处
《环境科技》
2025年第6期71-75,共5页
Environmental Science and Technology
基金
北京生态环境保护科学研究院科研项目(J2024-006).
关键词
多目标优化
减污降碳
协同
工艺优化
Multi-objective optimization
Pollution reduction and carbon reduction
Collaborative
Process optimization
