Enhancing offshore wind turbines performance with hybrid control strategies using neural networks and conventional controllers

Abstract Wind turbines are complex energy converters that require robust control solutions to address performance limitations caused by their mechatronic systems and external disturbances. Operating below the nominal power region implies significant challenges in achieving a precise, efficient response while maintaining structural stability and reducing vibrations. This study proposes a control architecture aimed at maximizing power generation and minimizing structural vibrations. Four hybrid control strategies are developed based on radial basis function neural networks (RBFNNs) combined with conventional regulators to address this dual objective. The proposed controllers compute the appropriate electromagnetic torque to track the maximum power point while mitigating tower acceleration. The RBFNNs utilize a non-supervised learning algorithm to adaptively adjust their weights, enabling better coupling with wind turbine dynamics. The hybrid control strategies were tested on a 5 MW floating offshore wind turbine subjected to the combined effects of wind and wave disturbances. Simulation results demonstrate that these methods achieve a more efficient power response while significantly reducing structural fatigue. The proposed hybrid strategies effectively enhance wind turbine performance by improving power generation efficiency and reducing mechanical stress, thereby extending the turbine's operational lifetime.