Investigation on the Multi-objective Optimization of Supercritical Airfoil based on Nondominated Sorting Genetic Algorithm

This paper aimed to investigate the multi-objective optimization method of supercritical airfoil. To achieve the optimal design of supercritical airfoil Rae2822, an improved NSGA-2 (Nondominated Sorting Genetic Algorithm) method was utilized, while the cross-operator and adaptive-variation operator were introduced to improve the convergence speed of the algorithm. During the optimization, the airfoil parametric modeling was achieved based on the Bezier-Bernstein method, and the objective function was obtained through solving the N-S equations. Considering the parallel computation characteristics of the algorithm, the computation was conducted in large-scale Linux computer system to reduce the solving time. Optimization results showed that the undominate solution with high quality obtained through the NSGA-2 method distributed evenly, which provided the designer a wider choosing space. It was also showed that the multi-objective optimization method presented in this paper was feasible and reliable.