基于多目标并行遗传算法的阻尼器优化布置

An Optimal Arrangement of Dampers Based on Multi-objective Parallel Genetic Algorithm

将快速非支配排序遗传算法和粗粒度-主从式并行遗传算法相结合,提出一种多目标并行遗传算法。该算法在同步优化消能减震结构上阻尼器的布置数量和布置位置时,既能多目标优化,又能并行运算,提高计算效率。采用该算法,以最大层间位移角和最大楼层加速度为多目标函数,优化一个14层钢框架结构上阻尼器的布置数量和布置位置。优化结果表明:该算法是逐渐收敛,具备较强收敛性;该算法优化结果适用于不同地震波,层间位移角减震系数最小为20.66%,加速度减震系数最小为18.35%;对于消能减震结构,综合考虑最大层间位移角和最大楼层加速度的影响时,建议阻尼器主要布置在结构中下部,且中部布置得要较多。

In this paper,a multi-objective parallel genetic algorithm is proposed by combining nondominated sorting genetic algorithm-Ⅱ and coarse-grained master-slave parallel genetic algorithms.When the algorithm is designed to optimize the number and position of dampers on an energy dissipation structure at the same time,it can be optimized by multiobjectives,and can be operated in parallel to improve the calculation efficiency.A 14-story steel frame structure is selected as the research object,and the optimal design of damper arrangement is conducted by taking maximal interstory drift ratio and maximal floor acceleration as the control objective.The results show the algorithm gradually converges and has a strong convergence.The optimized results of this algorithm are applicable to different seismic waves,the minimum damping coefficient of inter-story drift ratio and acceleration are 20.66% and 18.35%respectively.For the energy dissipation structures,the optimal locations of dampers are at the middle and lower floor and more in the middle floor when the maximal inter-story drift ratio and maximal floor acceleration are considered.