摘要
粒子群算法是一直典型的群体智能算法,种群多样性是影响其优化性能的一个重要因素,对几种影响种群多样性的方法如微粒的初始位置、迭代过程微粒个体执行变异操作、异步更新策略、带线性递减惯性权重的异步更新策略等进行研究.并通过对典型测试函数进行实验,给出了几种能提高微粒群算法优化性能的多样性控制方法.
Particle swarm optimization algorithm is a typical swarm intelligence algorithm, the diversity of the population is an important factor to affect the optimize performance of particle swarm optimization algorithm. Several methods that control the population diversity are analyzed, such as the initial position of the particle, the mutation of particle that perform during the iterative process , asynchronous update strategy, and asynchronous update strategy with a linearly decreasing i/aertia weight. Some test functions are used for simulation experiments, and several control methods of the diversity that can improve the performance of particle swarm optimization are given.
出处
《微电子学与计算机》
CSCD
北大核心
2013年第6期6-9,共4页
Microelectronics & Computer
基金
广西空间信息与测绘重点实验室开放基金项目(桂科能1103108-16)
关键词
粒子群算法
多样性
变异
异步更新
线性递减惯性权重
Particle Swarm Optimization（PSO）
diversity
mutation
Asynchronous update pattern
linearlydecreasing inertia weight
