New ways to control wind turbines using prediction models
Researchers reviewed different methods for regulating wind turbine power and motion.
Model Predictive Control (MPC) is used for wind turbine regulation. This method uses different internal prediction models to manage the machines.
Low-order linear or piecewise-affine predictors work for power and speed regulation. Flexible drivetrain, structural, and floating models are used for tower motion or blade loads.
Nonlinear and economic MPC studies show coordinated pitch and torque control. Preview, repetitive, and data-driven formulations address wind prediction and periodic loads.
Direct nonlinear applied-pitch propagation means the applied pitch is evaluated inside the original nonlinear aerodynamic map. This avoids using a linearized or learned surrogate.
An integration challenge remains at the intersection of flexible drivetrain prediction and nonlinear aerodynamic pitch treatment. This includes explicit constraints and validated model correction.
Key facts
- Low-order predictors are attractive for constrained power and speed regulation.
- Flexible models are needed when blade loads or platform dynamics are explicit control quantities.
- A remaining integration challenge exists for flexible drivetrain prediction and nonlinear aerodynamic pitch treatment.