Research on Carbon Emission Prediction Trend of Electric Power System Based on CS-SVR

Abstract:
In 2015, China pledged to the world that it would achieve carbon peaking by 2030 and strive to achieve carbon neutrality by 2060, and achieving power system emission reduction is a key step towards carbon neutrality in China. To respond to the call, as well as to provide theoretical support for China's goal of achieving peak carbon and carbon neutrality, a method for forecasting carbon emissions from China's power system based on the cuckoo algorithm (CS) and support vector regression (SVR) is proposed in this paper. The method used the Pearson correlation coefficient method to process the relevant influencing factors, and the influencing factors with high correlation coefficients, such as urbanisation rate, coal consumption in thermal power generation, electricity consumption in secondary industries and the amount of transmission and distribution losses, were used as inputs to the CS-SVR model, with carbon emissions as outputs. The results indicate that China will reach its carbon peak around 2030, and that the low carbon development trajectory is in a good position to achieve carbon neutrality by 2060.