This paper proposes a method for predicting unsteady aerodynamics of wind turbine airfoils using surrogate-based recurrence framework (SBRF) method. Using specified simulation results generated by the CFD method in some conditions, the unsteady aerodynamic model could be established by the Kriging surrogate model. Then, time-domain predictions of unsteady lift, moment, and drag in different conditions can be gained by the SBRF method with minimal computational expense. Some parameters have been set according to the operational condition of wind turbines so as to describe the unsteady aerodynamic modeling problem. The unsteady aerodynamic performance of the wind turbine airfoils in some training conditions is carried out by the commercial CFD simulator CFX, the results of which could be utilized to build the SBRF. Then the predicted time-varying aerodynamic characteristics of wind turbine airfoils in the validated condition could be obtained by the SBRF method and the CFD simulation, respectively. It is revealed from the results that the time-varying aerodynamic characteristics of wind turbine airfoils in most dynamic stall cases can accurately approximate by the SBRF method. In addition, the SBRF method has relatively less computational cost compared with the CFD method. Therefore, it can be used as the foundation of aero-elastic analysis and design optimization studies of wind turbines.
A robust methodology for estimating the aerodynamic characteristics of wind turbine airfoils based on errors in the outputs and in the inputs
