An Optimized Fractional Grey Prediction Model for Carbon Dioxide Emissions Forecasting

Area of Science:

• Environmental Science
• Data Science
• Mathematical Modeling

Background:

• Grey prediction models are useful for data not conforming to statistical distributions.
• The standard GM(1,1) grey prediction model has limitations in parameter determination and equal sample weighting.
• Accurate forecasting is crucial for environmental and economic policy, particularly for emissions data.

Purpose of the Study:

• To enhance the accuracy and applicability of grey prediction models.
• To address limitations in parameter estimation and data weighting within the GM(1,1) model.
• To develop a superior prediction model for carbon dioxide emissions using optimized grey prediction techniques.

Main Methods:

• Developed an optimized grey prediction model incorporating a genetic algorithm for parameter optimization.
• Implemented fractional-order accumulation to assign differential weights to sample data.
• Validated the model using International Energy Agency carbon dioxide emission data.

Main Results:

• The proposed optimized grey prediction model demonstrated significantly superior performance compared to existing models.
• The genetic algorithm effectively determined crucial model parameters, overcoming background value limitations.
• Fractional-order accumulation improved the model's ability to capture data regularities.

Conclusions:

• The optimized grey prediction model offers a more robust and accurate forecasting tool.
• This approach provides a significant advancement in grey system prediction methodologies.
• The findings have implications for more precise environmental emission monitoring and policy-making.