Machine learning and hybrid intelligence for wind energy optimization: A comprehensive state-of-the-art review

Wind energy plays a pivotal role in the global transition toward sustainable energy. However, its intermittent and stochastic nature presents challenges in achieving optimal performance, reliability, and seamless grid integration. Recent advances in machine intelligence—including machine learning (M...

Full description

Saved in:

Bibliographic Details
Published in	Expert systems with applications Vol. 296; p. 128926
Main Authors	Dubey, Ashutosh Kumar, Kumar, Abhishek, Segovia Ramírez, Isaac, García Márquez, Fausto Pedro
Format	Journal Article
Language	English
Published	Elsevier Ltd 15.01.2026
Subjects	Deep learning Fault detection Machine learning Multi-objective optimization Reinforcement learning Smart control Wind energy MPC DNN NMAE t-SNE CVSS Deep learning RANS Wind energy EEMD LASSO ANFIS RBM NARX WAFs D-GABC AI MAE RC RF RL DFIG RQ CFD SVD SVC CNN 5G-NG-RAN KNN SVM HHO SVR UAVs JVAE XGBR RLS GWO DCNN DQN SHAP LSTM ESS SOC WOA PRISMA REWS NRMSE PID VMD SOM Fault detection AD-PSO-WOA Machine learning LC CI/CD Multi-objective optimization MCP DRL RES DDQN LR ELM DL AUC QDA DT MLP BPNN DBN RNN FFT Meta-MARL GGWO EMD ML SCADA AGA-DDPG MQ NWP EMS MAPE MEC AEP PSO LUBE PCA WKNR NB GAN TCN Smart control NN TD3 MPPT LDA WPREs GRNN GMDH GRU IoT IGBT 2-DOF HARO AGA WT SSA TLBO EWT Reinforcement learning LiDAR GP RMSE RT2FNN RBFNN LIME WECs
Online Access	Get full text
ISSN	0957-4174
DOI	10.1016/j.eswa.2025.128926

Cover

More Information
Summary:	Wind energy plays a pivotal role in the global transition toward sustainable energy. However, its intermittent and stochastic nature presents challenges in achieving optimal performance, reliability, and seamless grid integration. Recent advances in machine intelligence—including machine learning (ML), deep learning (DL), and reinforcement learning (RL)—offer powerful tools to address these challenges across forecasting, control, maintenance, and diagnostics. This systematic review provides a comprehensive evaluation of how machine intelligence has contributed to the optimization of wind energy systems. These techniques have been applied to enhance turbine-level performance, reduce power losses, predict faults, and maximize energy yield under uncertain and dynamic conditions. Particular emphasis is placed on hybrid models that combine data-driven algorithms with physical dynamics and domain heuristics, enabling real-time, predictive, and autonomous wind farm operations. Furthermore, the study critically examines integration barriers such as noisy SCADA data, regulatory compliance, computational costs, and sustainability trade-offs. The findings highlight that multi-objective optimization—balancing energy production, system resilience, and cost efficiency—is central to the most successful implementations. Hybrid frameworks, explainable artificial intelligence (AI), edge computing, and transfer learning are identified as key enablers for scalable deployment. This review offers a comprehensive roadmap for the application of machine intelligence in advancing wind energy optimization and provides actionable insights for researchers, engineers, and policymakers committed to developing intelligent, adaptive, and sustainable wind power infrastructures.
ISSN:	0957-4174
DOI:	10.1016/j.eswa.2025.128926