摘要
针对最初的解析目标分流(ATC)法存在只适用于层次型多学科设计优化(MDO)问题和设计目标只能在第一级设置的问题,在非层次型MDO问题的每个子系统都引入设计目标,利用分解协调策略,提出了一种改进的解析目标分流(IATC)法。IATC是一种两级MDO方法,通过交替求解系统级和子系统级问题达到子系统间的一致性。通过两个典型解析算例的计算结果,验证了该方法的有效性,并且发现随着步长加大,误差变大、计算花费变小。
The previous analytical target cascading (ATC) method is fixed to hierarchical multidisciplinary design optimization (MDO) problems and the design targets can be only set at the first level. Design targets are introduced at each sub-system of non-hierarchical MDO problems and decomposition and coordination strategy is used, an improved analytical target cascading (IATC) method is presented. IATC is a two level MDO method, and the Consistency among all sub-systems is achieved through the system level problem and sub-system level problems being solved alternately. Two typical analytical test problems are adopted to verify the effectiveness of this method and a phenomenon is found that error gets larger and computation cost gets lower along with the step size increases.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第21期4701-4703,4731,共4页
Computer Engineering and Design
基金
国家863高技术研究发展计划基金项目(2006AA04Z119)
关键词
复杂系统
多学科设计优化
解析目标分流法
分解协调
二次罚函数
complex system
MDO
analytical target cascading
decomposition and coordination
quadratic penalty function