A fatigue-aware model predictive control framework for load mitigation in wind turbines during primary frequency regulation

Primary Frequency Regulation (PFR) requires wind turbines to rapidly adjust active power output, which inevitably introduces additional mechanical stresses and creates a fundamental conflict between grid frequency support and structural fatigue management. Existing control strategies typically employ fixed-gain parameters and lack explicit incorporation of dynamic fatigue constraints into the closed-loop design. To address this gap, this paper proposes a Fatigue-Aware Model Predictive Control (FAPC) framework for wind turbines(WTs) during PFR. A discrete-time state-space model is first developed to predict one-step-ahead variations of key fatigue indicators, specifically low-speed shaft (LSS) torque and tower lateral bending moment, thereby avoiding computationally intensive rainflow counting in real-time operation. Based on these predictive indicators, a multi-objective optimization scheme is formulated to dynamically coordinate the PFR gains for generator torque and pitch angle. A distinctive contribution of this work is the introduction of a phase-dependent active damping mechanism, which identifies the phase relationship between frequency deviations and structural vibrations and adaptively adjusts the control gains to dissipate structural energy while providing grid support. Simulation results demonstrate that the proposed framework achieves effective load decoupling: fatigue loads on the drive-train and tower lateral modes are reduced by 14.44% and 8.96%, respectively, while aerodynamic loads (tower fore-aft and blade bending) remain at baseline levels. Moreover, the maximum frequency deviation is reduced by 3.81%, confirming that structural protection is achieved without compromising, and even improving, frequency regulation performance compared to conventional fixed-gain methods.