Mathematical modeling and fuzzy logic control of horizontal axis wind turbine pitch drive system

In this thesis, an alternative Fuzzy-PI controller based on aerodynamic sensitivity (dP/dθ) for wind turbine systems is designed, replacing traditional PI controllers. While existing literature commonly adjusts gains based on rotor speed or fixed operating regions, this study proposes a control structure that is sensitive to the turbine's aerodynamic behavior. The developed method involves two Sugeno-type Fuzzy Inference Systems (FIS) that take error (e), change in error (de), and aerodynamic sensitivity (dP/dθ) as inputs, and calculate K_p and K_i gains separately. The proposed fuzzy controller was tested in the Simulink environment using the NREL 5MW reference wind turbine model and compared with the classical PI controller. The simulation results demonstrated that the proposed method reduced oscillations, improved transient state responses, and provided more stable rotor speed control. Furthermore, it was observed that this control structure is more sensitive and adaptive to wind speed changes. The Fuzzy-PI and S-Fuzzy-PI controllers achieved less oscillation and faster stabilization compared to the PI controller, particularly under rapidly changing wind speeds. This study demonstrates that aerodynamic sensitivity based gain scheduling is a feasible and effective approach in wind turbine control systems and emphasizes the advantages of a control strategy that is sensitive to the turbine's aerodynamic behavior. The results suggest that Fuzzy-PI controllers can be preferred, especially in large turbines, to ensure more efficient and reliable operation of turbine systems.