Prediction of CO2 emissions in China by generalized regression neural network optimized with fruit fly optimization...

Area of Science:

• Environmental Science
• Climate Change Research
• Computational Modeling

Background:

• Global warming necessitates urgent carbon emission reduction to meet China's carbon peak targets.
• Accurate carbon emission prediction is crucial for developing effective reduction strategies.
• Existing prediction models require enhancement for improved accuracy and reliability.

Purpose of the Study:

• To develop and validate a comprehensive model for predicting carbon emissions in China.
• To identify key factors influencing carbon emissions.
• To forecast future carbon emission trends and inform policy decisions.

Main Methods:

• Feature selection using Grey Relational Analysis (GRA) to identify significant emission drivers.
• Parameter optimization of Generalized Regression Neural Network (GRNN) using the Fruit Fly Optimization Algorithm (FOA).
• Forecasting carbon emissions from 2020-2035 using the developed FOA-GRNN model and scenario analysis.

Main Results:

• Fossil energy consumption, population, urbanization rate, and GDP were identified as major drivers of carbon emissions.
• The FOA-GRNN model demonstrated superior prediction accuracy compared to standard GRNN and Back Propagation Neural Network (BPNN).
• The study provides a forecasted trend of carbon emissions in China for the period 2020-2035.

Conclusions:

• The integrated FOA-GRNN model offers a robust approach for accurate carbon emission prediction.
• Policy makers can utilize these findings to set targeted emission reduction goals and implement effective energy-saving measures.
• This research supports China's efforts to achieve its carbon peak target amidst global climate change concerns.