An MPA-based optimized grey Bernoulli model for China's petroleum consumption forecasting.

Area of Science:

• Energy economics
• Time-series forecasting
• Mathematical modeling

Background:

• Petroleum consumption prediction is vital for energy scheduling and economic development.
• Uncertainty and volatility in petroleum systems necessitate advanced forecasting models.

Purpose of the Study:

• To develop a novel nonlinear grey Bernoulli model for accurate petroleum consumption forecasting.
• To enhance the model's predictive power using a combined fractional accumulated generation operator and marine predation algorithm.

Main Methods:

• Introduction of a combined fractional accumulated generation operator (incorporating traditional and conformable fractional accumulation).
• Application of the marine predation algorithm to optimize model coefficients, including fractional accumulation order.
• Verification of the model using coal consumption data before application to petroleum consumption.

Main Results:

• The proposed optimized nonlinear grey Bernoulli model (ONGBM(1,1,k,c)) demonstrates superior performance compared to benchmark models.
• The enhanced accumulation operator effectively extracts developmental patterns from time-series data.
• Accurate forecasts for China's petroleum and terminal consumption were generated.

Conclusions:

• The ONGBM(1,1,k,c) model offers a robust and accurate approach for petroleum consumption forecasting.
• The study provides valuable insights and suggestions for policymakers in the energy sector.
• Optimized fractional accumulation and meta-heuristic algorithms significantly improve time-series prediction accuracy.