Application Research On Carbon Dioxide Emission Prediction Model Of Urban Residential Buildings In Municipalities

In recent decades,China ’s economy has developed rapidly and has become the world’s second largest economy.The industrialization and urbanization process with coal as the main energy has made China the world ’s largest carbon dioxide emitter after 2008.The construction industry plays a vital role in achieving China ’s ’ carbon peak ’ goal by 2030.Therefore,if the Chinese government wants to reduce the total amount of carbon emissions,the reduction of carbon emissions in the construction industry is obviously an important part of it,and the total amount of carbon dioxide emissions has become the main measure of carbon emissions.In order to realize the low-carbon development of the construction industry,accurately measure and predict the total carbon dioxide emissions of buildings,determine quantitative emission reduction targets,and accurately formulate various emission reduction measures to achieve the predetermined energy conservation and emission reduction targets.Therefore,it is of great significance to accurately measure the carbon dioxide emissions of buildings and improve the accuracy of the carbon dioxide emission prediction model for the low-carbon development of buildings.However,due to the numerous driving factors of building carbon dioxide emissions,traditional prediction methods are difficult to achieve high-precision prediction of building carbon dioxide emission trends.In recent years,the combined optimization prediction model in machine learning has been widely used in many fields of prediction research due to its excellent prediction performance.By comprehensively comparing the existing research methods of building carbon dioxide emissions prediction,this paper uses the combined optimization prediction model in machine learning to predict and analyze the carbon dioxide emissions of urban residential buildings in Beijing,Shanghai,Tianjin and Chongqing.The specific research contents are as follows :Firstly,this paper analyzes the driving factors of carbon dioxide emissions for urban residential buildings in four municipalities in China.Through the literature analysis method,the primary index system of carbon dioxide emissions from urban residential buildings in four cities is established from the perspectives of demographic factors,economic factors and technical factors.In order to further ensure that the indicators in the established index system are highly correlated with the carbon dioxide emissions data of urban residential buildings,this paper is based on the random forest algorithm.The index system is further screened,and finally the carbon dioxide emissions prediction index system of urban residential buildings in four municipalities is selected.Secondly,aiming at the shortcomings of traditional prediction models in the prediction of building carbon dioxide emissions,this paper first proposes two prediction models of single model and optimization model in machine learning.After data experiments,the prediction results of three models,including traditional decision tree model,are compared and analyzed by error index.The results show that the error index of the traditional decision tree prediction model is greater than that of the other two models,and the prediction results deviate from the actual value seriously,while the stability of the other two single models is poor,and the prediction results need to be optimized.After the model is optimized,the optimization model is compared with the single model again,and the optimization model fitting diagram is closer to the actual value than the single model.In order to further improve the prediction accuracy,this paper uses Light GBM model as the base learner of Bagging ensemble learning strategy,and uses Grey Wolf Optimizer(GWO)to optimize the parameters of Light GBM model.Data experiments show that the established Bagging-GWO-Light GBM combined optimization prediction model is more suitable for the prediction of carbon dioxide emissions from urban residential buildings,and the prediction accuracy is higher.Finally,based on the established Bagging-GWO-Light GBM combined optimization prediction model,this paper predicts the carbon dioxide emissions of urban residential buildings in four municipalities in China from 2021 to 2025,and compares and analyzes the current situation and future development trends.In order to provide a scientific basis for the government ’s emission reduction work,it also provides a quantitative prediction analysis method for the prediction of carbon dioxide emissions from urban residential buildings in China.