Modeling and Estimation of CO₂ Emissions in China Based on Artificial Intelligence

Area of Science:

• Environmental Science
• Climate Change Research
• Socioeconomic Modeling

Background:

• China's rapid economic development post-reform has led to increased carbon dioxide (CO2) emissions.
• China has committed to carbon peaking by 2030 and carbon neutrality by 2060.
• Accurate CO2 emissions forecasting is vital for achieving these climate goals.

Purpose of the Study:

• To develop a hybrid intelligent algorithm for accurate CO2 emissions prediction in China.
• To support China's policy-making for carbon peaking and neutrality targets.

Main Methods:

• Developed a hybrid model combining Least Squares Support Vector Regression (LSSVR) and the Adaptive Artificial Bee Colony (AABC) algorithm.
• Optimized LSSVR hyperparameters using the AABC algorithm.
• Utilized socioeconomic indicator data from 1971 to 2017 for model training and validation.

Main Results:

• The developed hybrid model demonstrates high accuracy and robustness in CO2 emissions prediction.
• Achieved a relative error of approximately ±5% in advance CO2 emissions forecasting.
• The model provides reliable predictions for future CO2 emission trends.

Conclusions:

• The proposed hybrid intelligent algorithm offers a robust and accurate method for CO2 emissions forecasting.
• This modeling approach provides essential support for governmental and industrial climate policy formulation.
• The methodology can be adapted for forecasting other socioeconomic-related issues.