云环境下的多学科设计优化研究  

Multidisciplinary Design Optimization in Cloud Environment

作　　者：周际锋 姚锡凡[1] 刘二辉[1] 

机构地区：[1]华南理工大学机械与汽车工程学院,广东广州510641

出　　处：《机械设计与制造》2018年第2期153-156,共4页Machinery Design & Manufacture

基　　金：国家自然科学基金项目(51175187;51375168);汽车仿真与国家重点实验室开放基金资助(20121112);广东省交通运输厅科技项目(科技-2015-02-075)

摘　　要：多学科设计优化(MDO)在实际应用中面临计算量庞大而复杂的困难,而云计算为解决这种困难提供了手段。利用云计算整合优化资源方面的优势去解决多学科优化过程的难题,按云服务理念,将MDO作为一种服务,搭建了面向MDO的云优化平台,以目标级联法作为MDO的典型代表,对MDO问题进行层级分解,将复杂的工程系统分解为相互联系的若干子系统,降低了原优化问题的求解困难,以几何规划问题和汽车燃烧室的优化设计为例,探讨了云环境下MDO求解的有效性。Multidisciplinary design optimization （MDO） faces a diffwulty in large and complex computation in practical applications. Cloud computing provides an approach to solve such a difficulty. By means of the advantages of cloud computing in integrating the optimized resources to solve the difficult issues in the multidiseiplirmry optimization process, and according to the cloud service concept, this study treats MDO as a kind of service, establishes a cloud optimization platform for MDO, and addresses the validity of MDO in cloud environment by taking analytical target cascading as a typical MDO method, hierarchically decomposing problems of MDO, which decomposes the complex engineering system into several interconnected subsystems and reduces the solve difficulty of original problems, and by setting geometric programming problems and the optimization design of combustion chambers as examples.

关 键 词：云计算 多学科设计优化 目标级联法 云平台 

分 类 号：TH16[机械工程—机械制造及自动化] TP391[自动化与计算机技术—计算机应用技术]