摘要
针对粒子群算法快速收敛,容易陷入局部最优的现象,提出了一种以度为规则的小世界粒子群算法.该算法将粒子群的分布引入小世界模型,给小世界的随机连接赋予一个规则,即将度最大的粒子和度最小的粒子进行连接,使粒子之间充分交流,避免了粒子之间信息交流不充分而出现的早熟收敛现象.通过测试函数对其进行仿真实验,实验结果表明,以度为规则的小世界粒子群算法不易陷入局部最优,能够更好地寻找最优解,其性能优于传统的粒子群算法.
For fast convergence and prone to premature phenomenon of particle swarm optimization, proposed a small-world particle swarm optimization with a Rule of degree. The particle swarm distributions use the model of small world network. Give a rule for small-world＇s random connection, the biggest degree connect to the smallest, making communication more effective between particles, avoiding rapid convergence and has a balance of diversity of particle swarm. Using the standard function to test it, the experimental results show that the particle swarm not easy to fall into local optimum and better finding optimal solution. A small-world particle swarm optimization with a rule of degree is better than the conventional particle swarm algorithm.
出处
《微电子学与计算机》
CSCD
北大核心
2016年第10期115-119,共5页
Microelectronics & Computer
基金
国家自然科学基金项目(61379080)
山西省自然基金项目(2013011017-7)
关键词
粒子群
小世界
度
多样性
particle swarm
small-world
degree
diversity
