一种局部与全局相结合的微粒群优化算法  

Local and Global Combined Particle Swarm Optimization Algorithm

作　　者：胡乃平[1] 宋世芳[1] 

机构地区：[1]青岛科技大学信息科学技术学院,青岛266042

出　　处：《计算机工程》2008年第17期205-207,210,共4页Computer Engineering

基　　金：山东省自然科学基金资助项目(Y2005G02)

摘　　要：提出一种基于局部与全局搜索相结合的粒子群算法。该算法结合全局和局部PSO算法的优点,摆脱局部极优点的束缚,保持前期搜索速度快的特性,提高全局搜索能力。仿真实验表明,与标准微粒群优化算法相比,该算法的全局收敛性能得到显著提高,有效地避免微粒群优化算法中的局部收敛问题,并快速搜索到全局最优解。This paper proposes a particle swarm optimization based on local and global combined search. Combining the strongpoint of local search and global search, the new algorithm reduces the possibility of trapping at the local optimum. The advanced algorithm maintains the characteristic of fast search in the early convergence phase, and improves the global search ability. Experimental results indicate.that Local and Global Combined Particle Swarm Optimization （LGCPSO） has the advantage of convergence property over PSO. The algorithm avoids the local convergence problem effectively, and has the ability to get a convergence velocity quickly.

关 键 词：微粒群优化 局部 全局 仿真 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]