摘要
针对蚁群优算法在进化中容易出现早熟和停滞的现象,对基本蚁群算法进行了改进。借鉴生物群体的相互协作机理,将蚁群算法中的蚂蚁分成两个群体分别独立进行进化,并定期进行信息交换。同时,将遗传算法中排序的概念扩展到精英机制当中,形成基于优化排序的精英蚁群系统。两方法相结合,有效缓解了因信息素浓度失衡而造成的局部收敛,改进算法的搜索性能,计算结果也表明该算法有效性和可行性。
For the default that ant colony algorithm (ACA) presents the phenomenon of precocity and stagnation during evolution, the basic ACA is improved. Referred to the cooperation strategy in biosphere-colony, the algorithm separate the ants into two populations which evolves separately and exchanges information timely. An excellence mechanism of the ant colony system is also formed based on the optimize compositor. These methods can prevent local convergence caused by misbalance of the pheromone and can improve the searching performance of the algorithm effectively.
出处
《自动化技术与应用》
2008年第2期8-12,共5页
Techniques of Automation and Applications
基金
国防基础科研项目(9140A16070106CB0101)
关键词
蚁群算法
双种群
精英机制
排序
ant colony algorithm
dual population
excellence mechanism
compositor