Forecast Of Carbon Emissions In China Based On PSO-KGM（1,N）Model

Accurate carbon emission forecasts can provide a reference for the government to formulate emission reduction policies and play a pivotal role in slowing global warming.In this thesis,a multivariate nonlinear grey model based on kernel method（KGM（1,N））is applied for the first time to predict CO2 emissions,however,the uncertain parameters of KGM（1,N）may affect the prediction accuracy,and it is difficult to ensure that the established model is the optimal model for national and regional carbon emission data.Model parameters are optimized to improve its predictive performance and model applicability.The carbon emission data is affected by many factors and has nonlinear characteristics,and the KGM（1,N）model can effectively deal with such nonlinear data;there is a non-negligible correlation between the influencing factors,so ordinary least squares（OLS）is not suitable for model parameter estimation,while KGM（1,N）uses the kernel method to solve the convex optimization problem,and its stability will be stronger;in addition,most of the existing studies predict carbon emissions in one step.In order to perform multi-step prediction,this thesis adopts Grey relational analysis with timedelay effect to determine the influencing factors.The purpose of this thesis is to study the applicability and accuracy of the PSO-KGM（1,N）model in predicting CO2 emissions across the country and provinces.This attempt can provide some reference for related research and supplement the prediction method.Finally,the CO2 emissions of China and 30 provinces,municipalities and autonomous regions（except Hong Kong,Macao,Taiwan and Tibet）from 2021 to 2025 are predicted to provide reliable forecast results.The research results show that the proposed PSO-KGM（1,N）model has good prediction performance in the field of carbon emissions,and also has better accuracy and applicability compared with other methods.However,the predictive performance of the model decreases slightly with the length of the forecast period,so it is only suitable for short-term forecasts.According to the forecast results,Beijing is the only region with negative carbon emission growth,and it is expected to reach the peak of carbon emissions before 2025.Most of the provinces,municipalities and autonomous regions with poor emission reduction effect are located in the economically underdeveloped western and northeastern regions.If effective measures are not taken,China will not be able to achieve the carbon peaking target by 2030,and the regions will also be unable to achieve the proposed first-to-peak target.