Ensemble Kalman Filter Parameter Estimation for Individual Turbines

Genevieve M. Starke*, Carl R. Shapiro, Charles Meneveau, Dennice F. Gayme, Department of Mechanical Engineering, Johns Hopkins University


As wind farms become a more prevalent, wind farms need to be able to provide generator services such as frequency regulation. To provide a controller capable of frequency regulation, a reduced order model is used in the controller, which is tested on a large eddy simulation (LES) wind farm. This model is combined with an Ensemble Kalman filter to estimate the power production of the LES wind farm. The Ensemble Kalman filter incorporates power measurements from the wind farm to estimate the parameters determining the wakes of the turbines in the wake model. Using these new parameters, the wake model is able to follow the power production of each of the individual turbines in the farm.