基于自适应代理模型的气动优化方法

气动优化设计中，为了减少优化系统的计算周期，提高搜索效率，引入结构简单、计算量较小的代理模型，而运用有效的插值和选样方法(自适应选样)可以大大减少建立代理模型的时间。因此本文提出了一种基于自适应代理模型的气动优化方法。首先对自适应代理模型进行研究，建立了 Kriging 自适应代理模型和支持向量回归自适应代理模型，这两种自适应代理模型在相同样本点情况下比一般代理模型拥有更高的预测能力，然后将这其应用到翼型优化设计中，取得了良好的优化效果，从而表明这两种自适应代理模型不仅简单实用，而且明显提高了气动分析的计算效率。

Aerodynamic optimization method based on adaptive surrogate model

In order to reduce the computation cycle and improve the search efficiency,the Surrogate Model method with simple structure and low cost of calculation is applied in the aerodynamic optimization design. The combination of effective interpolation and sampling methods (adaptive sampling ) has been proved to be an effective reduction of the model establishing time.An aerodynamic optimization method based on the adaptive surrogate model is proposed in this paper.Based on the research of the adaptive surrogate model,the Kriging adaptive surrogate model and the support vector regression adaptive surrogate model are established.The models have higher predictive ability than the general surrogate model with same sample points. When applying to the airfoil optimization design, these two adaptive surrogate models perform well and satisfied aerodynamic optimal results are obtained,indicating that these two adaptive agent models are not only simple and practical,but also significantly improve the computational efficiency of aerodynamic analysis.