摘要
针对粒子群优化(PSO)算法随着维数增加而导致的收敛速度慢,容易陷入局部最优的问题,提出了一种合作式粒子群(CPSO)算法。通过多粒子群不同的组态向量合作,显著改善了标准算法的早熟问题。利用标准测试函数对CPSO算法、协同进化遗传算法(CCGA)、遗传算法(GA)、PSO算法进行比较测试,结果表明,CPSO算法在多个基准优化问题方面显示了较佳性能。
Particle Swarm Optimization(PSO) algorithm converges more slowly as the dimension increases, which easily cause the local optimum. A Cooperative Particle Swarm Optimization(CPSO) algorithm is presented, by using the way of cooperation to improve the premature convergence problem of standard algorithm. CPSO, Cooperative Coevolutionary Genetic Algorithm(CCGA), Genetic Algorithm(GA), PSO algorithm are compared with the test of standard function. The results indicate that CPSO algorithm shows better performance in a number of benchmark optimization problems, compared with the traditional algorithm.
出处
《太赫兹科学与电子信息学报》
2016年第2期276-281,共6页
Journal of Terahertz Science and Electronic Information Technology
关键词
收敛行为
合作式算法
合作种群
粒子群优化算法
convergence behavior
cooperative algorithm
cooperative population
Particle Swarm Optimization algorithm