结合极值优化的多粒子群协同进化算法  

A Multi-swarm Co-evolutionparticle Swarm Optimization Algorithm Combined with the Extremal Optimization Algorithm

作　　者：葛澎[1] 刘宏义[2] 

机构地区：[1]海军装备部驻西安军事代表局,陕西西安710054 [2]解放军边防学院,陕西西安710108

出　　处：《微电子学与计算机》2014年第6期176-179,共4页Microelectronics & Computer

摘　　要：多粒子群协同进化算法是一种群智能算法,具有智能性、通用性、并行性和全局搜索能力,能够很好地解决全局寻优问题,但其保持粒子多样性的机制和协同进化的机制有待做进一步的改进.为了进一步提高多粒子群协同进化算法的寻优效率,提出了一种结合极值优化的多粒子群协同进化算法,它将多粒子群协同进化算法的全局搜索能力与极值优化算法的局部搜索能力进行了结合.最后通过实验验证了该算法的有效性.The multi-swarm co-evolution particle swarm optimization algorithm is an algorithm based on swarm intelligence,with intelligence,versatility,parallelism and global searching ability,can solve the problem of global optimization very well,however,mechanism to keep the diversity of particles and co-evolution mechanism need to be further improved.In order to further improve the optimization efficiency of the muti-swarm co-evolution particle swarm optimization algorithm,we propose a new muti-swarm co-evolution particle swarm optimization algorithm combined with the extreme optimization algorithm(MCPSOEO),combine the global searching ability of the multiswarm co-evolution particle swarm optimization algorithm and the local searching ability of the extreme optimization. Finally,the experimental results prove the effectiveness of the algorithm.

关 键 词：多粒子群协同进化 极值优化 全局寻优 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]