Airfoil Optimization Design of Vertical-Axis Wind Turbine Based on Kriging Surrogate Model and MIGA

The aerodynamic optimization of the airfoil of vertical-axis wind turbines (VAWTs) is limited by the time-consuming nature of computational fluid dynamics (CFD), resulting in difficulty in the efficient implementation of multi-parameter optimization. In response to this challenge, this study constructed a collaborative optimization framework based on the Kriging surrogate model and the multi-island genetic algorithm (MIGA). Based on the NACA 0015 airfoil, 13 geometric variables (including 12 Bernstein polynomial coefficients and 1 installation angle) were defined through the Classification and Shape Transformation (CST) parameterization method. Through sensitivity analysis, seven key parameters were screened as design variables. Seventy training samples and ten validation samples were generated via Latin hypercube sampling to construct a high-precision Kriging surrogate model (R2 = 0.91368). The optimized results show that the power coefficient of the new airfoil increases by 14.2% under the condition of the tip velocity ratio (TSR > 1.5), and the average efficiency of the entire working condition increases by 9.8%. The drag reduction mechanism is revealed through pressure cloud maps and velocity field analysis. The area of the high-pressure zone at the leading edge decreases by 23%, and the flow separation phenomenon at the trailing edge is significantly weakened. This research provides an engineering solution that takes into account both computational efficiency and optimization accuracy for the VAWT airfoil design.