基于自适应免疫记忆克隆算法的空间网架结构减震装置优化配置  

Optimal allocation of damping devices in space grid structure based on adaptive immune memory clonal algorithm

作　　者：展猛 王社良[2] 张丽珍 ZHAN Meng;WANG Sheliang;ZHANG Lizhen(College of Arehiteeture Engineering Huanghuai University,Zhumadian 463000,China;College of Civil Engineering Xi'an University of Arehiteeture and Technology,Xi'an 710055,China;College of Mechanical and Energy Engineering Huanghuai University,Zhumadian 463000,China)

机构地区：[1]黄淮学院建筑工程学院,河南驻马店463000 [2]西安建筑科技大学土木工程学院,陕西西安710055 [3]黄淮学院机械与能源工程学院,河南驻马店463000

出　　处：《地震工程与工程振动》2018年第5期153-160,共8页Earthquake Engineering and Engineering Dynamics

基　　金：国家自然科学基金项目(51678480);河南省高等学校重点科研项目(19A560016)~~

摘　　要：针对遗传算法容易陷入早熟收敛和群体多样性差的问题,基于生物免疫系统中的克隆选择、免疫记忆以及免疫自调节机理,采用混沌化初始抗体群,并引入抗体浓度和抗体激励度,提出一种可以动态调节变异概率和克隆规模的自适应免疫记忆克隆算法(Adaptive immune memory clone algorithm,AIMCA)。以模态可控度作为优化目标准则的影响因素,分别采用改进的遗传算法(IGA)和AIMCA对一个85节点、288杆件的空间平板网架结构的作动器进行优化配置研究,分析两种算法在搜索能力、种群多样性和收敛速度上的差异,探讨AIMCA在作动器优化问题中优势和可行性。结果表明,遗传算法即使通过一定的改进,也很难满足对复杂问题的高维寻优要求,而AIMCA由于引入了记忆单元和免疫自调节机制,大大提高了其空间搜索能力,维持了较好的种群多样性;且收敛到同一精度,AIMCA收敛速度要比IGA快的多。在减震装置优化控制中,AIMCA可以获得更优的布置位置和更大的减震效果。The genetic algorithm was easy to fall into premature convergence and hadpoor diversity of population question;therefore,based on the clonal selection,immune memory and the immune self-adjusting mechanism in biological immune system,the paper put forward a kind of adaptive immune memory clonal algorithm (Adaptive immune memory clone algorithm,AIMCA)which can dynamically adjust the mutation probability and clone scale.The modal controllable degree was taken as the influence factor of the optimization criteria,and the improved genetic algorithm(IGA)and AIMCA were respectively used to discuss the optimization configuration of damping devices in aspace grid structure which has 85 nodes and 288bars.The difference of two algorithms in searching ability,population diversity and convergence speedwere analyzed;the advantage and feasibility of AIMCA for the optimization of damping devices was explored.Results show that although the genetic algorithmhas certain improvement,it is difficult to meet the requirements of optimization for complex problems;but because of the introduction of memory unit and adaptive immune mechanisms,AIMCA can greatly improve its space search ability and maintain the good population diversity.Converging to the same precision,the convergence speed of AIMCA is much more quickly than IGA'.In optimal control of damping devices,AIMCA can obtain better location and bigger damping effect.

关 键 词：遗传算法 免疫记忆 克隆算法 网架结构 优化配置 

分 类 号：TU311[建筑科学—结构工程]