基于量子粒子群算法的门式刚架结构抗风优化

为弥补目前结构抗风优化仅针对高层建筑的不足,采用量子粒子群算法对一大跨屋盖结构进行了抗风优化.基于风洞试验数据库获得等效静力风荷载,并根据型钢表组成离散变量搜索空间.通过约束违反协调系数,构造了一种新的适应值模型,进一步建立了粒子越界处理方法,以保证优化的可行性和收敛性.通过10次运行计算以确定门式刚架的最优设计,并在全风向角下对优化结果进行校核.研究结果表明,目标函数随迭代单调递减收敛,总质量标准差仅为其平均值的4%,平均迭代24次,说明量子粒子群算法用于门式刚架抗风优化具有较好的健壮性和计算效率. 

Wind-Resistant Optimization of Portal Frames Based on Quantum-Behaved Particle Swarm Algorithm

In order to make up for the insufficiency of researches on the structural wind-resistant optimization of large span roofs, the quantum-behaved particle swarm algorithm was adopted to optimize a large-span roof under wind loads. Equivalent static wind loads on the structure were obtained from the database of wind tunnel tests, and the search space of discrete variables was formed according to the shape steel table. An improved fitness function was proposed by defining a coordinate factor for constraint violation, and the over-flow dealing technique was constructed to guarantee the feasibility and convergence of the optimization. Ten runs of computation were carried out to determine the optimal design of the portal frame, and the validity of the optimal solution was checked in all wind directions. The research results show that the objective function decreases monotonically with respect to iterations, standard deviation for total mass is only 4% of its mean value, and the mean iteration number for every run is 24. Therefore, the quantum-behaved particle swarm algorithm displays strong robustness and high efficiency in the wind-resistant optimization of portal frames.