Research on dynamic prediction of carbon reduction potential in iron and steel parks

The steel industry is the main source of industrial carbon emissions, and accurate prediction is crucial to achieve the ‘ double carbon ‘ goal. However, data missing and hysteresis affect the prediction effect. In this paper, a dynamic prediction method based on power data is proposed. By analyzing the energy consumption of each part of the park, the relationship between power use and carbon emissions is explored, so as to realize the accounting and prediction of carbon emissions in the park. Firstly, the electric-carbon coupling model is established to determine the main emission sources. At the same time, the correlation coefficient is calculated by combining the gated cyclic unit and multiple regression. Finally, the long-term and shortterm memory network is used to predict the energy-carbon conversion coefficient, and the carbon emission prediction is realized by grey correlation analysis. Taking an iron and steel park in southern China as an example, the results show that the prediction accuracy is more than $\mathbf{9 5 \%}$.