Intelligent Control for Floating Offshore Wind Turbines Based on Game Theory with Data-Driven and Reliability Awareness

To address the challenges of complex modeling and control objective conflicts, a game control scheme based on data-driven and reliability awareness is proposed for floating offshore wind turbines (FOWTs). In this scheme, first, an enhanced model-free adaptive control method with the integration of iterative learning is proposed, which utilized online data for driving modeling and control command derivation to ensure the functionality of the FOWT and maintain stable power generation. Then, a healthy reliability-aware model of the pitching system is developed and integrated into the controller of FOWTs. Finally, a trade-off between the reliability and functionality of FOWTs is intelligently made through bargaining game theory to derive reasonable control commands. Simulation experiments of the proposed control scheme are carried out using FAST under different wind conditions. The results show that, compared with the gain-scheduled PI controller and model-free adaptive controller, the output power of the proposed controller is more stable and the pitch motion of the FOWT is significantly reduced; compared with the improved model-free adaptive controller, the proposed controller improves the reliability of the pitch system.