基于交叉因子和模拟退火的群搜索优化算法  

Integrated cross-factor and metropolis rule group search optimization algorithm

作　　者：杨文璐[1] 宁玉富[1,2] 

机构地区：[1]山东师范大学管理科学与工程学院,山东济南250014 [2]德州学院不确定系统实验室,山东德州253023

出　　处：《计算机工程与设计》2013年第6期2020-2024,共5页Computer Engineering and Design

基　　金：山东省科技发展计划基金项目(2009GG20001029)

摘　　要：针对目前标准群搜索优化(GSO)算法存在的一些缺点,提出一种基于交叉因子和模拟退火群搜索优化(CMG-SO)算法,通过与模拟退火算法的结合来改善算法的收敛性能,并借鉴遗传算法中的选择交叉操作增加粒子多样性,通过引入交叉因子增强群体成员优良特性,减小了算法陷入局部极值的可能。经过4个常用测试函数测试及与粒子群优化(PSO)算法、群搜索优化(GSO)算法对比,表明了该算法有较好的全局搜索能力和收敛速度,提高了优化性能。For some of the shortcomings of the current standard group search optimization （GSO） algorithm, proposed integrated cross-factor and metropolis rule group search optimization （CMGSO） algorithm, with the combination of the metropolis rule to improve the convergence of the algorithm performance, and drawing on the choice of crossover operation in genetic algorithm to increase particle diversity, enhancing the excellent characteristics of the group members introduce cross-factor to reduce the algo-rithm into a local extremum possible. Test CMGSO, particle swam optimization （PSO） and group search optimization （GSO） by the four commonly used tested function to find that this algorithm is better than these optimization algorithm at improving the global search ability, convergence speed and the optimized performance.

关 键 词：群搜索优化(GSO)算法 遗传算法 模拟退火算法 交叉因子 粒子群优化(PSO)算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]