基于竞选进化的群搜索优化算法_林敏.doc

基于竞选进化的群搜索优化算法 林敏1 张湘伟2 刘锋3 速云中1 (1. 广东工贸职业技术学院测绘遥感信息工程系 广州 510510； 2. 广东工业大学机械工程学院 广州 510006； 3. 广东工业大学土木与交通学院 广州 510006) 摘要：群搜索算法(Group Search Optimizer , GSO)是基于动物的群居行为特性建立的一种优化算法，并且较为成功地应用到连续变量的结构优化设计中，是一种新的群智能优化算法，适宜于解决多极值高维度优化问题，但其在优化的后期由于种群多样性不够，容易陷入局部最优。对GSO算法进行了改进，将竞选算法(Election Campaign Algorithm, ECA)的模型运用到GSO算法中，并引入群体适应度方差的概念来判断是否进行影响函数操作以提高收敛效率。将该算法与遗传算法(GA)、粒子群算法(PSO) 和基本的群搜索算法(GSO)进行优化测试函数的对比实验，并将其运用于建筑空间结构优化问题的研究，由于竞选算法(ECA)与群搜索算法(GSO)都具有群体社会性的共同特点，所以两者结合后，最终结果均验证了改进算法的有效性。 关键词：空间结构优化 群搜索算法 竞选算法 群体适应度方差 中图分类号：TU392.3 Campaign Group Search Optimization Algorithm Based On Evolutionary LIN Min1 ZHANG Xiangwei 2 LI Lijuan 2 LIU Feng 2 SU Yunzhong 1 (1. Remote sensing and Information EngineeringDepartment，Guangdong College of Industry and Commerce, Guangzhou 510510; 2. School of Mechanical Engineering, Guangdong University of Technology, Guangzhou 510006) Abstract：Group search algorithm (Group Search Optimizer, GSO) is an optimization algorithm based on the characteristics of animal social behavior, and successfully applied to the continuous variable structure optimization design, is a new swarm intelligence optimization algorithm, is suitable for solving multi extreme high dimensional optimization problems, but the optimization late because of population diversity is not enough, easy to fall into local optimum. The GSO algorithm is improved, will the election algorithm model is applied to GSO algorithm, and introduce the concept of the variance of the populations fitness to determine whether the influence function to improve the convergence efficiency of operation. The algorithm and genetic algorithm (GA), particle swarm optimization (PSO) and basic GSO algorithm are compared experimental optimization test functions, which will be used in the construction of spatial structure optimization, Since the election algorithm (ECA) and the group search algorithm (GSO) are the common characteristics of social groups, so the final