摘要
将遗传算法(GA)的全局寻优性能好和模拟退火算法(SA)的局部搜索能力强的优点相结合,提出了用于钢桁架结构离散变量优化设计的遗传模拟退火算法(SAGA).以十杆桁架为例对此算法进行了数值实验,并将实验结果与其他优化方法相比较.算例结果表明,遗传模拟退火算法的寻优概率是100%,平均进化代数为35代,其稳定性和求解效率均高于改进的遗传算法.实验结果显示,遗传模拟退火算法在整体搜索同时,采用退火操作进行局部搜索,提高了算法的局部搜索能力,有效克服了遗传算法迭代缓慢的缺点,把遗传模拟退火算法用于钢桁架离散变量的优化设计中是行之有效的.
To combine Genetic Algorithm（GA） with Simulated Annealing Algorithm（SA） that the Genetic Simulated Annealing Algorithm（SAGA） was proposed.It had the global searching ability of GA together with the local fast converging ability of SA.It was applied to the steel truss structural optimization with discrete variables and this paper provided the comparison between SAGA experiments and other optimal results.The experiments showed that the searching optimization probability of SAGA was 100% and the average evolved generations is 35,which indicated that SAGA was more stable and had better seeking efficiency than improved GA.The SAGA improved the local searching ability and overcame evolution slowness defect of GA through applying local searching annealing method.The SAGA is an effective method to seek the optimal design of steel trusses with discrete variables.
出处
《郑州大学学报（工学版）》
CAS
北大核心
2011年第6期54-57,共4页
Journal of Zhengzhou University（Engineering Science）
基金
国家自然科学基金资助项目(50378007)
关键词
遗传算法
模拟退火算法
优化设计
genetic algorithm
simulated annealing algorithm
optimal design
