Study On The Prediction Of Industrial Carbon Emissions In Jiangxi Province Based On IPSO-BP Neural Network Model

The continuous promotion of commercialization has brought about rapid growth of economy,which also leads to the gradual increase of carbon emission from fossil energy,thus causing the global problem of global warming.China is now in the middle of industrialization,and taking the new road of industrialization is the inevitable path at present.Since reform and liberalization,Jiangxi Province has been developing rapidly in the industrial economy,but high emissions and high pollution have always been obstacles on the road to green industrial development in Jiangxi Province,and the industrial environment has shown a backward technology,low energy efficiency utilization,and high industrial carbon emissions in the province’s share of the carbon footprint.At the same time,in order to cope with the "double carbon" target,Jiangxi Province should also take the initiative to undertake the development of low carbon economy,green promotion of key enterprises,the optimization of industrial restructuring,and increase the percentage of clean energy.This paper identifies the research object as industrial carbon emissions in Jiangxi Province.Firstly,the current situation of carbon emission research and carbon emission related theories at home and abroad are introduced,and GDIM influence factor decomposition method,SVR,PSO algorithm and BP neural network are analyzed.Secondly,based on the industrial end-use energy consumption data of 27 energy types in Jiangxi Province from 1996 to 2020,the carbon emission coefficient method is used to measure the industrial carbon emissions in Jiangxi Province and the results are analyzed.Again,based on the GDIM influence factor decomposition method to factor decomposition of industrial energy carbon emissions in Jiangxi Province,the model decomposes the influence factors of industrial energy carbon emissions in Jiangxi Province into eight dimensions: industrial gross product,total amount of industrial energy consumption,output carbon intensity,energy consumption carbon intensity,industrial employees,industrial per capita carbon emissions,industrial GDP per capita and energy consumption intensity,and conducts quantitative analysis.According to the degree of influence of the contribution rate of each influencing factor,six of them are selected as the input factors for carbon emission prediction.Finally,the SVR,BP and IPSO-BP prediction models were constructed,and the IPSO-BP neural network model combined with scenario analysis was selected to forecast the industrial carbon emissions of Jiangxi Province from 2021 to 2040 through fitting and comparison.Under the low-carbon and base-carbon scenarios,it shows a trend of first increasing and then slowly decreasing,and reaches a peak around 2035,while under the highcarbon scenario,it shows a trend of continuous growth.Based on the prediction results,targeted measures are proposed for energy conservation and reduction of industrial emissions in Jiangxi Province to provide some reference for the development of lowcarbon economy in Jiangxi Province.