基于改进引力搜索算法的桁架结构优化设计-《计算机技术与发展》

文章信息/Info

Title:: Optimization Design of Truss Structure Based on Improved Gravitational Search Algorithm

Author(s):: LI Yan-cang; YANG Zong-jin; Hebei University of Engineering,Handan 056038,China

Keywords:: test function; gravitational search algorithm; adaption; chaos mapping; struss optimization

摘要:: 引力搜索算法是近几年提出的较有竞争力的群智能优化算法,然而,标准引力搜索算法存在后期收敛速度慢的缺点。为有效利用优化算法来解决结构优化的问题,提出一种改进的引力搜索算法( improved gravitational search algorithm,IGSA)。通过引入 Logistic 映射,使 GSA 初始种群遍历整个搜索空间,提高算法找出最优解的可能性。通过引入粒子群算法( particle swarm optimization,PSO)的信息交互机制,利用个体粒子历史最佳位置和种群历史最佳位置动态调整粒子的速度和位置,使个体粒子更快地向适应度值更高的位置移动,使算法搜索能力加强。对 6 个经典测试函数进行寻优,结果表明改进后算法收敛速度快,收敛精度高,稳定性较佳,跳出局部最佳解的能力较强。用 IGSA 和 GSA 对 72 杆空间桁架进行尺寸优化,与其他算法相比,结果表明 IGSA 得到最优值的迭代次数明显减少,得到的最优解明显优于通用算法。

Abstract:: The gravitational search algorithm(GSA) is a competitive swarm intelligence optimization algorithm proposed in recent years.However,the standard gravitational search algorithm has the disadvantage of slow convergence in the later stage. In order to effectively solve the problem of structural optimization,an improved gravity search algorithm (IGSA) is proposed. By introducing Logistic mapping, the initial GSA population traverses the whole search space and improves the possibility of finding the optimal solution. Through the introduction of particle swarm optimization (PSO) information interaction mechanism, the particle speed and position can be dynamically adjusted by using the optimal position of individual particle history and the optimal position of population history,so that individual particles can move to the position with higher fitness value more quickly,and the search ability of the algorithm can be strengthened. Six classical test functions are optimized,and the results show that the improved algorithm has fast convergence speed,high convergence accuracy,high stability and strong ability to jump out of the local optimal solution. IGSA and GSA are used to optimize the size of the space truss of bar 72. Compared with other algorithms,the results show that the number of iterations of the optimal value obtained by IGSA is significantly reduced,and the optimal solution obtained is obviously superior to the general algorithm.