EGO算法的翼型气动外形优化设计

Airfoil shape optimization based on efficient global optimization

针对随机优化算法计算量大和最优响应面法容易陷入局部最优的缺点,采用EI最优策略综合平衡响应预测值及预测精度,建立了高效的优化系统.使用该方法进行了翼型气动外形优化设计,结果表明该方法将翼型阻力系数降低22%,具有良好的优化精度,而总计算耗时与粒子群算法相比约降低68%,说明了该方法的可行性和有效性.

To solve the problem that the random optimization method costs huge calculation and the best response method tends to get a weak local optimization, an effective optimization algorithm is built in this paper, which use best expected improvement（EI） strategy to balance response value and response precision. Tests of airfoil shape optimization indicate that the drag coefficient is reduced by 22% and the calculation time is reduced by 68% comparing with that of the particle swarm optimization, which shows that this effective optimization algorithm is realizable and effective.