Optimized ANFIS-Based Control of Variable-Speed Wind Turbines Using Bayesian Optimization for Enhanced Efficiency and Adaptability

Wind turbines with variable speed operation are crucial for maximizing energy efficiency and reducing mechanical stress in fluctuating wind conditions. This study enhances the Adaptive Neuro-Fuzzy Inference System (ANFIS) for controlling such turbines by introducing a novel hyperparameter tuning approach using Bayesian Optimization. The proposed method dynamically explores the hyperparameter space to optimize the fuzzy membership functions and training parameters, significantly improving model accuracy and reducing computational costs. MATLAB simulations demonstrate that the optimized ANFIS achieves higher power output and smoother control under variable wind speeds compared to conventional methods. The results show a notable improvement in turbine efficiency, a reduction in mean squared error (MSE) by 15%, and enhanced system adaptability to real-time wind variations. These findings underscore the potential of combining Bayesian Optimization with ANFIS for advanced wind energy management systems.