摘要
全面学习微粒群优化算法使用所有其它粒子的历史最好信息来更新粒子速度的策略改进标准微粒群算法,虽然一定程度避免陷入早熟,然而也存在到算法后期收敛速度急剧变慢的问题。采取兼顾粒子搜索范围和收敛速度方法并引入自适应的策略监视算法过程,当算法陷入停滞,即重新初始化,更新粒子系统,用复杂组合测试函数进行测试,表明了改进算法的有效性。
Comprehensive learning particle swarm optimizer (CLPSO) is studied, which uses a learning strategy whereby all other particles' historical best information to update a particle's velocity, has improved the standard PSO and has avoided premature convergence to some extent. However, during the late stages of algorithm, the convergence speed has deeply declined. Particle's searching scope and convergence speed are taken into consideration and the self-adapting strategy is used to monitor the process of algorithm, once it comes to a standstill, the whole particle's system is reinitialized and updated. The effectiveness is demonstrated through several experiments that are performed using composition test functions.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第8期1963-1965,1968,共4页
Computer Engineering and Design
基金
江苏技术师范学院青年科研基金项目(KYY06081)