一种用于结构拓扑优化的改进遗传演化算法  

Improved Genetic Evolutionary Algorithm for Structural Topology Optimization

作　　者：肖雨果 盛鹰[1] ZENG Lingxi 

机构地区：[1]四川大学建筑与环境学院,成都610065 [2]Risk Management Institute,National University of Singapore,singapore119077

出　　处：《四川理工学院学报（自然科学版）》2016年第2期36-45,共10页Journal of Sichuan University of Science & Engineering(Natural Science Edition)

基　　金：国家自然科学基金委员会与中国工程物理研究院联合基金资助(U1430119)

摘　　要：拓扑优化是结构优化设计领域中的重点和热点课题。结构渐进优化法(ESO)和遗传演化算法(GESO)是近年来提出的拓扑优化方法,可用于寻找结构最优的拓扑形状,指导结构概念设计。但这两种方法均存在诸多缺点,在工程应用中具有局限性。针对ESO和GESO方法的缺陷,发展了适用于拓扑优化的一种基于并行小生境比对法的改进遗传算法,并将改进遗传算法与ESO方法相结合,提出了灵敏度过滤技术和单元删除的修正判据,建立了引导式单元删除策略和孤立单元判断机制,形成了一种新的改进遗传演化算法(Improved Genetic Evolutionary Algorithm,IGEA),并以VC++作为编程平台,编写了集改进遗传算法、有限元、IGEA方法于一体的拓扑优化程序,有效实现了结构拓扑优化设计。算例表明,提出的IGEA方法优于ESO和GESO方法。Topology optimization is the hot and difficult research topic in structure optimization. The evolutionary structural optimization（ ESO） and genetic evolutionary structural optimization（ GESO） are both topology optimization methods proposed in recent years,and can be used to search for optimal topology and guide conceptual design of structure. However,both of the methods have many disadvantages,thus have limitations in engineering applications. In accordance with the disadvantages of ESO and GESO,an improved genetic evolutionary algorithm（ IGEA） by using improved genetic algorithm（ GA） and ESO was proposed in this paper and could be applied in structural topology optimization effectively. Many improvements such as parallel niches comparison method,filtering technology of element sensitivity,the modified criterion and guiding strategy of element deletion and judgment mechanism of isolated elements were proposed in the new algorithm.With the programming platform of VC＋＋,the topology optimization program of the improved GA,finite element method and IGEA was completed. The numerical examples show that the proposed IGEA method is better than ESO and GESO methods.

关 键 词：拓扑优化 改进遗传演化算法 灵敏度过滤 引导式单元删除 孤立单元判断 

分 类 号：TU311[建筑科学—结构工程]