【作者】 单志辉； 刘学军； 吕宏强；

【Author】 Shan Zhihui~1,Liu Xuejun~1,Lv Hongqiang~2 (1.College of Information Science and Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China) (2.College of Aerospace Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

【机构】 南京航空航天大学信息科学与技术学院； 南京航空航天大学航空宇航学院；

【摘要】 翼型设计过程中需要反复评估翼型气动性能,传统的风洞试验和计算流体力学(CFD)评估方法成本高、速度慢。本文在翼型气动性能评估中引入高斯过程回归模型,以辅助传统评估方法。选取一组NACA系列翼型,计算出每个翼型的升力系数和阻力系数;利用其中部分翼型的数据训练模型,寻找翼型外形与气动性能的对应关系;再利用训练好的模型预测其余翼型的气动性能。通过对模型的预测准确度、残差分布、训练集大小、推广性能和计算效率的研究表明,高斯过程回归模型在翼型气动性能评估中是非常有效的,尤其是大大快于传统的气动性能评估方法。更多还原

【Abstract】 Airfoil aerodynamic performance evaluation plays an important role in airfoil design.However, traditional wind tunnel tests and computational fluid dynamics(CFD) methods are costly and computationally slow.In this article,a Gaussian process regression model is introduced to aid the evaluation of airfoil aerodynamic performance using the traditional methods.The aerodynamic performance of a group of NACA series airfoils is calculated.Some of the resulting data are used to train the Gaussian process regression model,trying to find the hidden relationship between airfoil shape and airfoil aerodynamic performance.We then use the trained model to predict the aerodynamic performance of the remaining airfoils.Study of the prediction accuracy,residual distribution,training set size,generalization and computational efficiency indicates that the Gaussian process regression model is highly valid on the evalualion of airfoil aerodynamic performance.Particularly,it is much faster compared to either of the traditional ways.更多还原