基于小生境遗传算法的离散变量结构优化设计  

Structure Optimization of Discrete Variables Based on Niche Genetic Algorithm

作　　者：叶青[1] 韩英仕[1] 谢军[1] 

机构地区：[1]辽宁工学院土木建筑系,辽宁锦州121001

出　　处：《辽宁工学院学报》2007年第1期29-32,35,共5页Journal of Liaoning Institute of Technology(Natural Science Edition)

摘　　要：小生境技术的引入,提高了遗传算法处理多峰函数优化问题的能力。提出了基于隔离机制的自适应小生境技术,隔离小生境技术具有生物学基础,不仅能够有效地保证群体中解的多样性,而且具有很强的引导进化能力,针对简单遗传算法中的交叉与变异概率等不能动态地适应整个寻优过程,提出采用根据适应度调整交叉、变异概率并与小生境技术相结合的改进遗传算法。算例表明,该遗传算法对桁架结构的布局进行优化设计比较容易实现,简单、有效,可以产生很好的效益。The adoption of niche concept improved the ability of genetic algorithm in solving multimodel function optimization problems. A new niche model which is based on the isolation mechanism and self-adaptiveness, was proposed. The isolation niche technique has its basis on biology, it not only can maintain the diversity, in population, but also has a strong function of guidance for evolution. Aiming at simple genetic algorithm with invariable crossover and mutation probability dynamically unfit for the whole optimal process, a modified genetic algorithm with adaptive crossover and mutation probability mixed, the niche concept was presented in this dissertation. The exemplification results expatiated that the method proposed is really simple and efficient to the layout optimization of truss structures.

关 键 词：隔离小生境遗传算法 结构优化 绗架 

分 类 号：TU311.4[建筑科学—结构工程]