基于SVR多学科设计优化代理模型技术研究

为挖掘设计潜力,提高飞机设计质量,通过代理模型方法研究了多学科设计优化(MDO)。基于径向基函数模型,研究了一种新的代理模型技术--支持向量回归方法(SVR)。通过数值算例和某通用航空飞机多学科设计优化应用算例对支持向量回归方法的有效性进行了验证。研究结果表明,支持向量回归是一种比较有效的代理模型方法,在飞机多学科设计优化的理论研究和工程实践方面具有重要的参考意义。

Study of surrogate model technology in multidisciplinary design optimization based on SVR

In order to exploit design potentialities and improve the quality of design,multidisciplinary design optimization(MDO)isstudied with the use of surrogate model method.Based on radial basis function(RBF)model,a new surrogate model technology,support vectors regression(SVR)is discussed.Numerical examples and a MDO application example of one general aviation aircraft are calculated so as to demonstrate the validity of SVR.It’s shown that SVR is a very effective method to build surrogate model and has strong reference for theoretical research and engineering practice in aircraft MDO.