A method for protecting a wind turbine from an extreme change in wind direction includes receiving a wind direction and/or a wind speed at the wind turbine. When a change in the wind direction or the wind speed exceeds a predetermined threshold, the method includes determining a margin to stall and/or zero lift of the at least one rotor blade of the wind turbine as a function of an angle of attack or change in the angle of attack at a blade span location of at least one rotor blade of the wind turbine. The method also includes implementing a corrective action for the wind turbine (without shutting down the wind turbine) when the margin to stall and/or zero lift exceeds a predetermined value so as to avoid stall and/or negative lift on the at least one rotor blade during operation of the wind turbine.

A method for operating a wind turbine includes determining at least one wind condition of the wind turbine for a plurality of time intervals. The method also includes determining a status of the wind turbine at the beginning of each of the plurality of time intervals. Further, the method includes determining at least one vibration parameter of the wind turbine for one or more preceding time intervals of the plurality of time intervals. Moreover, the method includes predicting whether a trip event is imminent based on the at least one wind condition, the status of the wind turbine at the beginning of each of the plurality of time intervals, and the vibration parameter. Thus, the method further includes implementing a control action for the wind turbine so as to prevent the trip event.

A method of parameterizing a controller of a first wind energy installation wherein the controller sets a manipulated variable of the wind energy installation as a function of an input variable. An artificial intelligence determines at least one value of a parameter of the controller for at least one state/degree of being iced up of the wind energy installation based on a power curve, load, and/or downstream flow of the wind energy installation predicted with a mathematical model of the wind energy installation for at least one state/degree of being iced up, and/or determines at least one value of a parameter of the controller for at least one state/degree of being iced up of the wind energy installation, based on at least one determined state/degree of being iced up and a power, load, and/or downstream flow of the wind energy installation and/or at least one second wind energy installation.

A wind turbine includes a rotor, a plurality of rotor blades coupled to the rotor, and a blade pitch control system coupled to each rotor blade. A computer-implemented method for controlling the wind turbine includes determining at least one pitch position for a first blade. The method also includes determining whether there is a malfunction of the blade pitch control system associated with the first blade. The method further includes predicting a rotor imbalance using a model of at least a portion of the wind turbine. The method also includes comparing the predicted rotor imbalance with a predetermined threshold value. The method further includes one of regulating the pitch position for the second blade such that the predicted rotor imbalance is restored to a value below the predetermined threshold and regulating a pitch position for a second blade such that the predicted rotor imbalance does not exceed the predetermined threshold.

Provided is a method of controlling at least one wind turbine in case of a rotational overspeed situation, the method including: determining a current state related to the wind turbine; providing data related to the current state as input to a turbine model; predicting a load of at least one wind turbine component and power output of the wind turbine using the turbine model provided with the input for plural control strategies; comparing the predicted load and power output for the plural control strategies; and selecting that control strategy among the plural control strategies that satisfies a target criterion including the load and the power output.

The present invention relates tocontrol of wind turbines based on predicted operational trajectories. A control system for a wind turbineis described wherea main controller calculating one or more predicted operational trajectories and a safety controller validates at least one of the one or more predicted operational trajectories. The control system controls the wind turbine with the predicted control trajectory if the validation is valid, and controls the wind turbine with a safe-mode control trajectory if the validation is invalid. In an embodiment, the main controller is implemented as a receding horizon controller, e.g. in the form of a model predictive controller (MPC).

The invention relates to techniques for verifying a nacelle yaw position sensor installed on a wind turbine and for taking restorative action to control the nacelle yaw position. The invention relates to a method performing the comprising determining a first absolute wind direction signal associated with the first wind turbine; determining a second absolute wind signal direction signal associated with the plurality of other wind turbines; comparing the two wind direction signals; and issuing a nacelle yaw position sensor fault signal if the first signal is beyond a predetermined error range of the second signal. A benefit of the invention is that it enables the detection of an inaccurate nacelle yaw sensor without direct measurement or inspection.

Method of operating a wind turbine comprising a plurality of blades rotatable along their longitudinal axes using a pitch mechanism, and comprising one or more movable trailing edge surfaces. The method includes predicting, at a first moment in time, a high load for one or more of the blades at a second moment in time. The method further comprises actuating on one or more of the movable trailing edge surfaces of these blades such that the trailing edge surfaces have a wider range of control to counteract the predicted high loads before the second moment in time, and simultaneously pitching the blades such as not to negatively affect the operation of the wind turbine. The method furthermore comprises, at the second moment in time, actuating the one or more movable trailing edge surfaces of the at least one or more blades to counteract the high loads.

Methods, systems, and devices for controlling a yaw offset of an upstream wind turbine based on reinforcement learning are provided. The method includes receiving data indicative of a current state of the first wind turbine and of a current state of a second wind turbine adjacent to the first wind turbine downstream along a wind direction, determining one or more controlling actions associated with the yaw offset of the first wind turbine based on the current state of the first wind turbine, the current state of the second wind turbine, and a reinforcement learning algorithm, and applying the determined one or more controlling actions to the first wind turbine.

Controlling a wind turbine including measuring a wind speed for a location upwind of a wind turbine. Using the measured wind speed, a changed steady-state deflection of a structure of the wind turbine is predicted. The predicted changed steady-state deflection corresponds to a time when wind from the location is incident on the wind turbine. Oscillations of the structure are damped relative to the changed steady-state deflection. By damping the oscillations relative to the changed steady-state deflection, movements of the structure may be minimized when there is no predicted change in steady-state deflection, while permitting more rapid movements during transitions from one steady-state deflection to the predicted steady-state deflection, allowing more of the available power to be captured by the wind turbine.