摘要
针对传统的人工蜂群算法具有较强的勘探能力,但是算法局部开采能力较弱,演化后期收敛速度慢,容易陷入局部最优,提出了基于混沌算法和逆向学习算法相结合的初始化种群方法,有效改进种群的多样性;提出了一种新型的搜索策略来改进观察峰与侦察蜂的搜索过程,加快了算法的收敛速度;通过5个标准函数进行测试,文中算法在搜索效率、最优解质量、稳定性均优于传统的人工蜂群算法.
Traditional artificial bee colony algorithm has a stronger exploration capability, but its local ex- ploitation capability is a little weak and convergence speed in the late stage of evolution is slow, liable to fall into local optimum. A method of initialization of population based on chaos algorithm and inverse study algorithm is put forward, which can improve the diversity of population. A novel search strategy is proposed to improve the search process of onlooker bees and scout bees so as to speed up convergence speed. The algorithm used in this paper is proved superior to the traditional algorithm in search efficiency, optimal solution quality and stability through the test of five standard functions.
出处
《许昌学院学报》
CAS
2013年第2期62-67,共6页
Journal of Xuchang University
基金
河南省重点科技攻关项目(122102210488)
许昌市科技攻关项目(1101029)
许昌学院青年骨干教师资助计划
关键词
人工蜂群算法
混沌算子
逆向算子
artificial bee colony algorithm
chaos operator
inverse operator
