Airfoil optimization using a machine learning-based optimization algorithm

Abstract For the design of wind turbines, airfoil optimization is widely required as operation efficiency turbines closely dependent on aerodynamic performance, where accuracy method great importance. In this paper, a machine learning-based algorithm proposed to improve performance. A low-speed NACA0012 selected original for optimization. The class-shape-transformation (CST) used construct geometry airfoil, and performance calculated using panel code XFOIL. validation simulation carried out by comparing predicted with experimental data. order maximizing lift-to-drag ratio main objective problem while maintaining lift coefficient not smaller than values. results show that present has fairly good convergence, can obtain much higher optimized compared one. Compared traditional genetic method, learning based achieve better shorter time same problem. future, very promising fluid machinery