邻域加强的并行子空间优化过程  

Neighborhood Strengthened Concurrent Subspace Optimization Based on Response Surface

作　　者：欧阳琦[1] 陈小前[1] 黄奕勇[1] 

机构地区：[1]国防科技大学航天与材料工程学院,湖南长沙410073

出　　处：《国防科技大学学报》2011年第5期22-25,共4页Journal of National University of Defense Technology

基　　金：国家自然科学基金资助项目(50975280);教育部新世纪优秀人才支持计划资助项目(NCET-08-0149)

摘　　要：针对基于响应面的并行子空间优化(CSSO-RS)过程的不足,提出了基于邻域加强的改进CSSO-RS优化过程。其主要思路是在子空间优化后,在得到的最优解附近进行全部变量的优化,以更好地协调系统级优化与子空间优化。用经典测试函数及通用航空飞机参数优化问题对该算法进行测试,测试结果表明该算法使CSSO-RS优化效率得到较大的改善。A neighborhood strengthened CSSO-RS is proposed to improve the disadvantages of the Concurrent Subspace Optimization based on Response Surface（CSSO-RS）.This algorithm gives rise to an optimization between the system level optimization and subspace level optimization,which greatly improves the coordination affection between these two levels.The optimization region of this algorithm is the neighborhood of the optimal solution obtained by the subspace optimization,and all the design variables are involved in this optimization.The classical test function and an aircraft concept sizing problem were tested,and the efficacy of the algorithm was proved.

关 键 词：并行子空间优化 响应面 邻域加强 飞机参数优化 

分 类 号：V221[航空宇航科学与技术—飞行器设计]