基于自动分组遗传算法的建筑结构体系与构件尺寸协同优化设计  

Concurrent optimization of structural systems and sizes based on automatic grouping genetic algorithms

作　　者：刘晓峰 程耿东[2] 

机构地区：[1]金风科技股份有限公司,北京100176 [2]工业装备结构分析国家重点实验室,大连理工大学工程力学系,大连116024

出　　处：《计算力学学报》2014年第1期1-7,共7页Chinese Journal of Computational Mechanics

基　　金：国家自然科学基金(90816025)资助项目

摘　　要：研究了利用结构优化方法实现建筑结构体系选择的问题。文中建立了可以覆盖框架、框架-剪力墙和巨型结构三种体系的优化模型,并采用自动分组遗传算法求解,实现了结构体系与构件尺寸的协同优化。文章特别设计了三类模块,每类模块通过专门构造的含有多个分量的设计变量来表示,按一定方式组合这些模块可得到不同体系的结构。巨型结构中,巨型构件的设计变量表示方法和惯性矩计算方法满足提出的三条假设。基于精心构造的设计变量和三条相关假设,文中建立了以结构总材料用量最少为目标,考虑强度、刚度、分组及构造要求等约束条件的优化模型,采用自动分组遗传算法研究了40层、10层、6层三种高度建筑的优化设计。在相同外荷载条件下,它们的最优设计分别为巨型结构、框架-剪力墙和框架结构。This paper studies the choice of structural systems based on the optimization method. An opti- mization model involving the structural systems of frame, framed tube and mega-structure is presented. The concurrent optimization of structural systems and sizes is achieved by an automatic grouping genetic algorithm（AGGA）. Three types of modules which may compose the different structures are designed specially, and each module is represented by an design variable with several sub-variables. Three assump- tions of mega-members relating to the representation of design variables and the computation of inertial moment are proposed. Based on the proposed design variables and three assumptions, a minimum total weight optimization model subject to constraints of strength, stiffness, grouping and construction is ob- tained. The proposed optimization model together with the AGGA is used to the design optimization of three structures with 40-stories, 10-stories and 6-stories respectively, and three different optimal struc- tures,i, e. mega-structure, framed tube structure and frame structure,are obtained.

关 键 词：自动分组 遗传算法 结构体系 优化 

分 类 号：O224[理学—运筹学与控制论]