摘要
提出了一种基于差分进化算子变异的改进微粒群优化算法,为减小陷入局优的可能性,在群体最优信息陷入停滞时引入差分进化算子变异,使算法摆脱局部极优点的束缚,同时又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,该文算法的全局收敛性能得到了显著提高,能有效避免微粒群优化算法中的早熟收敛问题。
This paper proposes a modified particle swarm optimization （MPSO） with differential evolution operator mutation. When the optimum information of the warm is stagnant, differential evolution operator mutation is introduced to reduce the possibility of trapping at the local optimum. By adding the mutation operator to the PSO algorithm, the advantaged algorithm can maintain the characteristic of fast speed in the early convergence phase and improve the global search ability. The experimental results indicate that MPSO not only has great advantage of convergence property over PSO, but also can avoid the premature convergence problem effectively.
出处
《计算机工程》
CAS
CSCD
北大核心
2006年第15期25-27,共3页
Computer Engineering
基金
国家"973"计划基金资助项目(2002CB3122000)
国家自然科学基金资助项目(60074027)
国家"863"计划基金资助项目(2003AA412010)
关键词
微粒群
优化
差分进化
变异
Particle swarm
Optimization
Differential evolution
Mutation
