Abstract: 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.

Key words: Particle Swarm Optimization(PSO), local, global, simulation

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。

关键词: 微粒群优化, 局部, 全局, 仿真

CLC Number: 

• TP301.6