| With the increase in energy demand in the transportation industry year by year,Peaking carbon emissions from the transportation industry as soon as possible plays a key role in China’s overall realization of the carbon peaking and carbon neutrality goals.At present,the research on carbon peaking in the transportation industry is still in the initial stage,and it is of certain research significance to reasonably and accurately predict the carbon emissions of transportation and explore the path of carbon peaking.By combing the research results of carbon emission estimation,influencing factors,prediction and peaking path in the transportation industry,combined with the theory of sustainable development and low-carbon transportation,and comprehensively used emission factor method,least absolute shrinkage and selection operator(LASSO),long short-term memory(LSTM)and scenario simulation to explore the the path to peak carbon emissions in the transportation industry.The main research contents are as follows.(1)In-depth analysis of the current status of energy consumption and development in my country’s transportation industry.On this basis,using the emission coefficient method to estimate the carbon emissions of my country’s transportation industry from 1990 to 2019.Sort out the factors affecting carbon emissions in the transportation industry from multiple angles,and use the LASSO model for variable screening The six indicators of economic level,urbanization level,energy structure,transportation structure,and transportation development level are important influencing factors for carbon emissions in the transportation industry.(2)Based on key factors affecting the transportation industry.The LSTM network is applied as the underlying predictive model.By integrating a sliding window,the input of the network is improved.To study the impacts of future emission reduction policies on carbon emissions of transportation industry,the scenario analysis based on dynamic policies is used to construct a prediction model.A polynomial error fitting method is used for error correction to improve the model accuracy.The selected data is segmented into subsequences by the sliding window.The prediction accuracy under different window lengths are compared to select the optimal window parameters for the improved LSTM model.The improved LSTM model obtained is then compared with the original LSTM,BPNN and RNN models.The prediction results verified the effectiveness of the method.(3)The development trend of carbon emissions in the transportation industry from 2020 to 2035 is predicted by setting four emission reduction scenarios: baseline scenario,clean scenario,structural transformation scenario,economic slowdown scenario,and low-carbon scenario.The research results show that in the benchmark situation,the transportation industry cannot achieve the goal of carbon peaks before 2030.Cleaning scenarios are more significant than the carbon reduction effect of structural transformation.Low-carbon scenario reached its peak in 2030,which is an effective path to achieve a carbon peak as soon as possible.Based on the basis of the above research,the corresponding policies and suggestions are put forward. |
