基于快速群搜索算法的双层球面网壳多目标优化  

Multi-objective Optimization of Double-layer Spherical Shells Based on Quick Group Search Algorithm

作　　者：金晶[1,2] 李丽娟[1] 刘锋[1] 何嘉年[1] 

机构地区：[1]广东工业大学土木与交通工程学院,广东广州510006 [2]广东省建筑设计研究院,广东广州510370

出　　处：《建筑科学与工程学报》2014年第4期97-102,共6页Journal of Architecture and Civil Engineering

基　　金：国家自然科学基金项目(51178121);广东省自然科学基金项目(S2012020011082)

摘　　要：提出一种新的群智能算法,即多目标快速群搜索优化(MQGSO)算法,以双层球面网壳为研究对象,进行结构节点最大位移最小化和结构质量最轻化的多目标优化设计,网壳优化后的Pareto前沿分布均匀且范围较广,分别对多目标优化和单目标优化后的结构进行多维地震作用时程响应分析。结果表明:多目标优化得到的结构优于单目标的优化结果,所提出的多目标快速群搜索算法能很好地实现复杂空间结构的多目标优化设计。The multi‐objective quick group search optimization （MQGSO） algorithm was applied to the optimal design of double‐layer spherical shells . The optimization objectives were the minimum value of the maximum displacement and the weight of the structure .Taking the double‐layer spherical shells as an object ,the Pareto frontier of the optimization example structure had the full and broader distribution .Corresponding to the multi‐objective optimization results and the single‐objective optimization results ,the time history response analysis of optimized structure under multi‐dimensional earthquake action was carried out . The results show that the multi‐objective optimization results are better than those of the single‐objective one . The dynamic characteristics of the double‐layer shell example prove that the M QGSO algorithm can effectively deal with the multi‐objective optimal design of complex spacial structures .

关 键 词：双层球面网壳 多目标优化 快速群搜索算法 时程分析 PARETO前沿 

分 类 号：TU318[建筑科学—结构工程]