Research on Carbon Emission Prediction Model Based on Power Load Data

In the context of the global active response to climate change, China has firmly put forward the “dual carbon” goals. However, there are many problems in the current field of carbon emission research. Traditional carbon emission prediction methods have low accuracy and struggle to comprehensively and accurately capture the nonlinear relationship between carbon emissions and numerous complex influencing factors, resulting in a large deviation between the prediction results and the actual situation; the stability is also not good. Facing external environmental changes or data fluctuations, the output results fluctuate significantly and cannot provide continuous and reliable data; in addition, it is difficult to obtain input data. Its influencing factors cover many fields such as economy, energy and industrial production. Data sources are widely distributed, and the collection, collation and integration work is challenging, limiting the accuracy and practicality of the prediction model. The power industry is a key area for carbon emissions. This paper focuses on studying the relationship between power data and carbon emissions, optimizes and improves the BP neural network, and builds an electricity-carbon prediction model for carbon emission prediction on this basis. Finally, through the data of a certain province, the proposed electricity-carbon prediction model is simulated, and the electricity-carbon prediction model constructed in this paper is comprehensively compared and deeply analyzed with other prediction models. The research results clearly show that the carbon emission prediction model proposed in this paper can effectively achieve carbon emission prediction in the power field.