Carbon Emission Forecast of Electric Power Industry Based on Lasso and IPSO-BP Neural Network Model

Abstract—In the development of low-carbon power, the calculation and prediction of carbon emissions in the power industry are fundamental tasks. In order to improve the prediction accuracy of carbon emissions in the power industry and achieve the energy-saving and emission reduction goals of the power sector, this paper uses the Lasso regression model to screen out five important carbon emissions influencing factors based on the panel data of the power industry from 2001 to 2020. The simulation setting of the value of each influencing factor from 2021-2035 is simulated setting. The IPSO-BP neural network model was established to predict the carbon emissions and peak time of the power industry from 20121-2035. The prediction results show that under the simulated scenario, the carbon emissions of the power industry will increase year by year from 2021 to 2029, and will reach a carbon peak of 482,937,500 tons in 2029, and then decline year by year.