并行子空间优化在无人机总体设计中的应用

飞机设计是一个多学科的复杂的系统工程,各个学科通常相互影响、相互耦合.这使得飞机设计过程日趋复杂,设计周期越来越长,开发成本越来越高,而并行子空间优化(CSSO)是解决这些问题的一种有效方法.文中对基于神经网络响应面的并行子空间优化算法及其在无人机总体方案设计优化中的应用进行了研究.并行子空间优化算法将多学科耦合的无人机设计优化问题分解为不同的子空间问题,在不同的子空间中建立各自的神经网络响应面,通过响应面完成各子空间之间的数据交换与协调,以此来逼近设计空间最优解.应用结果表明,CSSO算法能有效地应用于无人机总体方案优化设计.

Application of Concurrent Subspace Optimization in UAV Preliminary Design

Aircraft design is a complicated system engineering involving many disciplines.Typically,all disciplines are coupled and influence each other.As a result,aircraft design becomes more and more complicated,and the longer the design cycle is,the higher the development cost goes.The Concurrent Subspace Optimization(CSSO) is a kind of effective algorithm for solving such problem.The Concurrent Subspace Optimization based on the neural network response surface and the Application of CSSO in the scheme optimization design of UAV are described.The CSSO has resolved the multidisciplinary coupling optimization design problems of UAV through decomposing the system into different subspace,and the neural network response surface is constructed in different subspace.The data exchange and coordination among the subspaces could be completed by the approximation of the response surface.Using this method the optimal result of the design can be obtained.The results indicate that such an implementation of the CSSO in the scheme optimization design of UAV is quite encouraged.