Enhanced Framework for Wind Power Forecasting Yields Improved Accuracy

Strategic Insights -

Researchers from the College of Engineering have proposed an innovative framework for wind power forecasting that incorporates a data transformation mechanism with a multi-objective none-dominated sorting genetic algorithm III (NSGA-III) and a hybrid deep Recurrent Network (DRN) and Long Short-Term Memory (LSTM) architecture. This approach has been shown to significantly improve the accuracy of wind power forecasting, outperforming existing frameworks with an error reduction of up to 99.87% in terms of Mean Squared Error (MSE) and 98.73% in terms of Root Mean Squared Error (RMSE).

Key Takeaways:

  • The proposed framework integrates a data transformation mechanism with a multi-objective NSGA-III and a hybrid DRN-LSTM architecture for wind power forecasting.
  • The framework achieves an accuracy improvement of up to 99.87% in terms of MSE and 98.73% in terms of RMSE compared to existing frameworks.
  • The study identifies the optimal subset features from wind energy datasets using the feature selection algorithm NSGA-III.
  • The data transformation process is applied to the selected features before inputting them into the hybrid DRN-LSTM for wind power forecasting.
  • The hybrid DRN-LSTM architecture is shown to outperform existing frameworks with a comparative study demonstrating its superior effectiveness and robustness.
  • The study's contributions lie in its approach integration of data transformation mechanism and the notable enhancements in wind power forecasting accuracy.
  • The research team, led by Yahya Z. Alharthi, includes Haruna Chiroma, Lubna A. Gabralla, and Princess Nourah Bint Abdulrahman University, Riyadh, Saudi Arabia, as financial supporters.

Statistics:

  • The proposed framework achieves an accuracy improvement of up to 99.87% in terms of MSE (8.8814e-07 on the classical algorithm vs 2.6593e-10 on the proposed framework).
  • The study demonstrates a 98.73% reduction in RMSE (9.424e-04 on the classical algorithm vs 1.630e-05 on the proposed framework).
  • The data transformation process is applied to 80% of the selected features before inputting them into the hybrid DRN-LSTM.

Sources:

  • "Enhanced framework embedded with data transformation and multi-objective feature selection algorithm for forecasting wind power." Scientific Reports, 2025,15(1):1-20.
  • Information Technology Newsweekly. May 27, 2025; p 578.