基于遗传算法的建筑结构优化设计方法研究  

Research on optimization of structural design based on genetic algorithm

作　　者：邵晓根[1] 姜代红[1] 王雷[2] 

机构地区：[1]徐州工程学院信电工程学院,徐州221008 [2]徐州市公安局科技处,徐州221116

出　　处：《江苏科技大学学报（自然科学版）》2017年第6期821-824,共4页Journal of Jiangsu University of Science and Technology:Natural Science Edition

基　　金：国家住房和城乡建设部科学技术项目(2013-K2-5);江苏省建设系统科技项目(2014JH18);徐州市科技计划资助项目(KC15SH049)

摘　　要：针对传统遗传算法存在容易过早收敛、寻优效率较低、精度不高等缺点,从适应度值函数标定和群体多样化两方面对传统遗传算法进行了改进,避免了传统遗传算法过早陷入局部最优解,拓宽了寻优空间;将改进的遗传算法应用于建筑结构优化设计中,通过建立以质量最小为目标的优化数学模型,解决具有应力约束和截面尺寸约束的离散变量结构优化问题,并对改进型遗传算法进行优化设计结果比较;结果表明,改进型遗传算法演化代数低于标准遗传算法,收敛性能明显更佳,提高了遗传算法在结构优化应用方面的计算速度和优化效果.Traditional genetic algorithm has disadvantages such as pron to premature convergence,low efficiency optimization and precision etc. Both the fitness value function and diverse group of traditional genetic algorithm have been improved to avoid the premature falling into traditional genetic algorithm local optimal solution,broadening the optimization space. The improved genetic algorithm is applied to optimize the design of building structures,through the establishment of minimum weight objective optimization model to solve the cross-sectional dimension with stress constraints and the constraints of structural optimization with discrete variables,and then with standard genetic algorithm,the design of the results is optimized. The results show that the improved genetic algorithm needs fewer iterations than the standard genetic operators,the convergence performance is significantly improved,and calculation is sped up in structural optimization.

关 键 词：遗传算法 建筑结构 结构优化 全局最优 数学模型 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]