几何参数化建模是气动布局设计的关键技术之一，简洁、高效、准确的几何表征对于提高飞行器设计效率和质量有着至关重要的作用。基于主成分分析（PCA）的特征提取降维，可以在满足几何表征精度的条件下，进一步降低现有参数化方法的维度，更好地服务于气动布局设计。本文介绍了基于PCA的翼型参数建模方法，分析了采样空间设计、样本数量、采样参数、几何重构方式等各个因素对PCA建模过程中降维能力、基模态特性以及几何表征能力的影响。通过CFD仿真分析，进一步探究PCA建模方法在气动性能表征方面的能力。仿真分析指出：基于PCA的翼型参数化方法，可以在满足几何外形表征精度的条件下降低现有方法参数维度，且其设计参数与几何特性有对应关系，利于在布局设计中加入工程经验；基于同一类采样方式的PCA建模，其模态特性、降维能力和几何外形拟合能力受采样空间、样本数量的影响很小，但对采样方法的参数配置较为敏感；本文所研究的建模方法，可在保证几何表征精度的同时满足气动力表征精度，其在气动布局设计优化中，具有一定的指导意义。

Geometry parametrization plays a significant role in Aerodynamic Shape Optimization （ASO）. A succinct， accurate， and efficient geometric presentation method can effectively improve the optimization efficiency and the design results. Principal Component Analysis （PCA） is a common way to extract data features and reduce data dimension. In this paper， an airfoil geometry parameterization method based on the PCA is firstly introduced. Then， the influences of sampling space， sample number， sampling method and geometric reconstruction method on PCA dimension reduction ability， base mode characteristics and geometric presentation ability are analyzed. Furthermore， the presentation ability of the PCA method on aerodynamic characteristics is also studied based on computation fluid dynamics. Simulation results show that： the PCA method could effectively describe the airfoil geometry shape with preferable accuracy and relatively few， physically meaningful parameters； based on a specific sampling method， the PCA mode， the dimension reduction ability and the geometric accuracy are little affected by the sampling space and the sample number， though they are sensitive to the parameters of the sampling method； the PCA method described in this paper could not only accurately describe the geometry shape， but also ensure the accuracy of the aerodynamic forces to a certain degree， which has certain guiding significance in ASO applications.