期刊文献+

一种基于试验设计与近似模型的计算软件设计方案 被引量:1

Framework Design of Computational Software Based on Experimental Design and Approximate Model
原文传递
导出
摘要 响应面方法、径向基函数人工神经网络方法以及Kriging方法等基于试验设计的近似模型方法是工程结构可靠度计算的常用建模方法,在结构动力响应预测、复杂结构优化设计以及多学科优化等方面也得到了广泛应用。该文重点研究相关方法的工程软件设计方法,目的是实现软件体系结构的开放性和可扩展性。运用面向对象方法及UML语言设计了试验设计与近似模型拟合软件系统框架,提出了主要类的设计方案和接口定义。利用C++程序设计语言以及Qt软件包开发了相关程序模块,实现了4种常用试验设计方法与2种近似模型。该软件系统设计采用了软件设计模式中几种常用模式,在此框架基础上可以方便扩展新的试验设计方法和近似模型方法及相关模块。该软件系统进行了经典算例验证,并应用于结构可靠性分析;在对某工程结构的动力响应优化计算中,综合应用了该系统的试验设计和近似模型模块,并与遗传算法结合,获得了良好的优化效果。 Response surface method,radial basic function,artificial neural network,and kriging method.etc,which based on experimental design,are commonly employed in the reliability calculation.In addition,they are also widely applied to structural dynamic response prediction,optimal design of complex structures and MDO.This objective of framework design is to achieve system's openness and expansibility.The software framework of experimental design and approximate models has been designed using Object-Oriented methods and the UML language.Interfaces of main classes and function definitions are proposed.The software system is developed with C++ programming language and Qt package.Four experiments design methods and two approximate models are implemented.In the software design, several common software design patterns are employed;therefore the system facilitates integration of new experimental design approaches,approximate model methods and other new models.The software system has been extended to conduct structural reliability analysis and has been verified by some benchmarks.And then it is successfully applied to design optimization of dynamic response for an engineering structure.
出处 《武汉理工大学学报》 CAS CSCD 北大核心 2010年第9期271-274,共4页 Journal of Wuhan University of Technology
关键词 结构可靠度 近似模型 软件设计 动力响应 结构优化 structural reliability approximate models software design dynamic response structural optimization
  • 相关文献

参考文献6

二级参考文献61

共引文献173

同被引文献3

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部