Forecasting the wind energy using hybrid deep learning model

Kaysal K., Hocaoğlu F. O., ÖZTÜRK N.

Evolving Systems, cilt.16, sa.4, 2025 (SCI-Expanded, Scopus) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 16 Sayı: 4
  • Basım Tarihi: 2025
  • Doi Numarası: 10.1007/s12530-025-09756-z
  • Dergi Adı: Evolving Systems
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC
  • Anahtar Kelimeler: Artificial bee colony, CNN, Forecasting, LSTM, Wind power plant
  • Gazi Üniversitesi Adresli: Evet

Özet

Accelerating population growth and ongoing technological progress have markedly intensified the global demand for electrical energy. This increase has caused a shift towards renewable energy sources. Wind energy, in particular, offers an environmentally friendly solution. However, this energy creates challenges due to its natural variability and non-stationary characteristics driven by seasonal and geographical factors. Accurate short-term forecasts are essential to ensure grid reliability and energy planning. This study proposes a hybrid deep learning model that integrates CNN, LSTM and GRU models optimized through metaheuristic algorithms such as Artificial Bee Colony (ABC), Particle Swarm Optimization (PSO) and Genetic Algorithm (GA). The proposed model aims to increase the forecast accuracy by effectively capturing both spatial and temporal features of wind energy data. For this purpose, the CNN-LSTM model optimized with the ABC algorithm exhibited superior performance compared to other optimization methods with an R² value of 0.989.These findings not only increase the accuracy of the forecast, but also provide guidance for investors in strategic planning by allowing the estimation of electricity production from wind energy at the regional level. This study makes an important contribution to the literature with a hybrid method that supports the effective use of wind energy potential.