摘要
传统的遗传算法在处理离散变量结构优化问题存在很多缺陷 ,遗传进化过程缓慢 ,遗传操作具有相当大的盲目性 .本文采取改进措施 ,实施最差个体替换法和优良个体保护法 ,确保遗传种群的优良品质 ,保证优良个体只执行改进性能的基因操作 .同时利用拟满应力设计方法收敛快的特点 ,将拟满应力设计方法和遗传算法相结合提出了组合遗传算法 ,适时地利用拟满应力设计优化结果提供种群成员 ,从而加快遗传进化的进程 .算例表明 ,组合遗传算法的优化结果是令人满意的 .
The traditional genetic algorithm (GA) had many shortcomings with respect to the structure optimization with discrete variables. The iteration process was slow and genetic operator had file prospect of much randomization. An approach of replacement for the worst and protection for the best is present in this paper, which ensures the high quality of the population. The excellent individuals only execute the genetic operation which improves their performance. The imitative full stress design method (IFS) was given in reference [Ⅰ]. The iterative procedure of IFS converges fast. The combinatorial genetic algorithm combines IFS and GA, which accommodates the result of IFS in the population of GA. The superiority of each method can be exerted and shortcomings evaded. The result of example is very satisfied.
出处
《沈阳建筑工程学院学报》
2001年第3期183-186,共4页
Journal of Shenyang Archit Civil Eng Univ: Nat Sci
基金
辽宁省自然科学基金资助项目 ( 962 0 0 9)
关键词
离散变量
遗传算法
结构优化设计
discrete variables
IFS
GA
combinatorial genetic algorithm
