桁架形状优化的自适应协方差矩阵进化策略方法  

Shape optimization of truss using covariance matrix adaptation evolution strategy method

作　　者：苏国韶[1] 武振兴[1] 燕柳斌[1] 

机构地区：[1]广西大学土木建筑工程学院,广西南宁530004

出　　处：《空间结构》2011年第2期35-41,共7页Spatial Structures

基　　金：国家自然科学基金资助项目(50809017);中国博士后科学基金(20080440812)

摘　　要：自适应协方差矩阵进化策略(CMA-ES)是在进化策略算法(ES)基础上发展起来的一种新的全局优化算法,对于处理复杂非线性多峰值优化问题具有很好的适用性.结合有限元方法,提出一种桁架形状优化的自适应协方差矩阵进化策略方法.采用空间25杆桁架和平面37杆桥形桁架两个桁架形状优化的经典算例对方法的可行性和先进性进行验证,其中,空间25杆桁架分为不考虑局部稳定约束和考虑局部稳定约束两种情况进行计算.研究结果表明,该方法是可行的,与基于遗传算法、粒子群优化算法等现代全局优化算法的桁架形状优化方法相比较,具有寻优效率高、收敛速度快、全局优化能力强的优点,在获得相同精度最优解的条件下,调用有限元分析的次数明显较少,从而有效地减少了计算耗时.Covariance matrix adaptation evolution strategy algorithm （CMA-ES）, which is a newly developed global optimization algorithm based on classical evolution strategy algorithm, has.become a powerful tool for solving highly nonlinear optimization problems. Combining finite element method, a truss shape optimal method based CMA-ES algorithm is proposed. Two classical examples involving both 25 bar space truss and 37 bar plane bridge-like truss are conducted to prove the feasibility and efficiency of the proposed method. The conditions with local stabilization and without local stabilization are both considered in the example of 25 bar space truss. The results show that method is feasible. Compared with the shape optimization method based on genetic algorithm and particle swarm algorithm, the proposed method has the advantages of high efficiency, fast convergence rate and strong capacity of searching global optimal solution, and needs much less number of calling finite element code when solving problems of shape optimization of truss. As a result, the computation time of optimization calculations for shape of truss is reduced dramatically by use of the proposed method.

关 键 词：结构优化 形状优化 桁架 自适应协方差矩阵进化策略算法 

分 类 号：TU323.4[建筑科学—结构工程]