Skip to Main Content
Model predictive control (MPC) is receiving attention in wind turbine controller design due to its ability to explicitly handle state and input constraints. Robust model predictive control (RMPC) additionally accounts for uncertainty in the future evolution of the system. Here, RMPC is compared with nominal MPC for the purposes of fore-aft tower damping of large wind turbines. The two controllers are identical save for their handling of the multiplicative and additive uncertainty in the prediction horizon. The comparison is performed by means of fatigue analysis in a state-of-the-art aeroelastic simulation package. State and input constraints are applied to a control model that is identified by data-driven methods. The robust controller bounds the uncertainty with a sequence of polytopes, which tighten the constraints to reduce constraint violations, while retaining the computational complexity of a quadratic program.