Research on Enterprise Carbon Flow Deconstruction and Emission Reduction Optimization Technology Based on Electricity Big Data

This study, based on an electricity big data platform, conducts optimization analysis and dynamic prediction of carbon flow deconstruction and sectoral carbon emissions in Guangxi. By integrating multi-source data, a carbon-electricity coupling analysis system encompassing 600 million data points was constructed, highlighting the significant correlation between electricity consumption and carbon emissions in energy-intensive industries (e.g., metal smelting and chemical manufacturing), with correlation coefficients reaching as high as 0.86 in some sectors. Using the ARIMA model and trend extrapolation methods, the study predicted the trends in the industry's carbon-electricity transmission coefficients for 2021–2023. Results indicate an upward trend in high-emission industries such as non-metallic mineral manufacturing, while primary industries and certain manufacturing sectors show a decline. High-frequency monitoring results reveal seasonal fluctuation patterns in carbon emissions, with daily emissions for some large industrial enterprises reaching 20,398.95 tons, peaking in March and August. The findings provide scientific support for precise carbon monitoring and dynamic management in Guangxi, offering a robust data foundation and methodological framework for formulating effective carbon reduction strategies.