摘要
针对粒子群算法容易陷入局部最优的缺陷,在分析惯性因子在算法中的作用机理的基础上,设计了一个根据种群多样性和进化代数自适应调节的惯性因子,并运用试探法,通过变换搜索步长,提高算法的局部搜索能力。最后,给出了3个典型函数的模拟例子,通过与APSO的对比结果显示,改进后的算法其性能得到极大提高。
Aiming at premature defect and poor result of Particle Swarm Optimization algorithm, a new Self-adaptive inertia factor was designed according to diversity in the population and generation number based on analysing inertia factor＇ s effect of algorithm. And through ploughing around adjusting step factors, the Particle＇s ability in local searching was enhanced. Three typical function tests were given. Comparing with APSO, the result indicates the effectiveness of this improvement.
出处
《计算机科学》
CSCD
北大核心
2009年第11期193-195,共3页
Computer Science
基金
国家自然基金(No.F0975026)
陕西省自然科学研究计划项目(No.2007f19)资助
关键词
粒子群算法
惯性因子
进化代数
Particle swarm optimization algorithm,Inertia factor,Generation number
