Improved exponential smoothing grey-holt models for electricity price forecasting using whale optimization.

Area of Science:

• Energy Economics
• Computational Intelligence
• Time Series Forecasting

Background:

• Accurate electricity price forecasting is crucial for energy market stability and decision-making.
• Existing models often struggle to adapt to dynamic market conditions and evolving trends.
• The need for computationally efficient forecasting tools is paramount in the fast-paced energy sector.

Purpose of the Study:

• To introduce a novel Whale Optimization Algorithm (WOA)-based multivariate exponential smoothing Grey-Holt (GMHES) model for electricity price forecasting.
• To enhance the adaptive capabilities of forecasting models by optimizing parameters using WOA.
• To evaluate the performance and efficiency of the proposed WOA-GMHES(1,N) model on real-world electricity price data.

Main Methods:

• Development of the WOA-GMHES(1,N) model integrating WOA for adaptive parameter optimization.
• Utilizing historical electricity price data to capture underlying trends and market dynamics.
• Evaluation of the model's accuracy and computational efficiency using RMSE and SMAPE metrics on Cameroonian electricity price data.

Main Results:

• The WOA-GMHES(1,N) model demonstrated superior performance compared to competing models.
• Achieved high accuracy with RMSE of 12.63 and SMAPE of 0.01%.
• Exhibited computational efficiency, generating forecasts in under 1.3 seconds.

Conclusions:

• The WOA-GMHES(1,N) model provides a robust and accurate solution for electricity price forecasting.
• The adaptive nature of WOA enhances the model's ability to capture evolving market dynamics.
• The model's efficiency and accuracy make it a valuable tool for time-sensitive energy sector decisions.