Repetitive model predictive approach to individual pitch control of wind turbines

Wind turbines are inherently exposed to nonuniform wind fields with of wind shear, tower shadow, and possible wake contributions. Asymmetrical aerodynamic rotor loads are a consequence of such periodic, repetitive wind disturbances experienced by the blades. A controller may estimate and use this peculiar disturbance pattern to better attenuate loads and regulate power by controlling the blade pitch angles individually.

A novel model predictive (MPC) approach for individual pitch control of wind turbines is proposed in this paper. A repetitive wind disturbance model is incorporated into the MPC prediction. As a consequence, individual pitch feed-forward control action is generated by the controller, taking ”future” wind disturbance into account. Information about the estimated wind spatial distribution one blade experience can be used in the prediction model to better control the next passing blade. A simulation comparison betweeen the proposed controller and an industry-standard PID controller shows better mitigation of drive-train, blade and tower loads.