Research on China's Carbon Emission Peak and Trend Based on Support Vector Machines and Improved BP Neural Network Models

Research on China's Carbon Emission Peak and Trend Based on Support Vector Machines and Improved BP Neural Network Models

연구 분야: Networking

학회: 2024 5th International Conference on Computer Engineering and Application (ICCEA)

This paper focuses on the prediction and analysis of China's carbon emission peaks and trends through intelligent algorithms, in order to provide a scientific and realistic basis for the policy formulation of economic development and environmental protection. Firstly, a particle swarm algorithm optimised support vector machine prediction model (PSO-SVM) is designed and constructed to improve the computational speed and accuracy of prediction through the iteration of particle swarm algorithm, so as to predict the trend of carbon emission in China. Secondly, the improved neural network model using genetic algorithm (GA) and principal component analysis (PCA) was constructed, and the average error was trained to less than 5% after iterative training, so as to predict the carbon emissions. Finally, the prediction results of the two intelligent algorithms are compared. The results show that China's carbon peaks around 2035, after which China's carbon emissions will show a decreasing trend year by year.