摘要
为解决蚁群聚类易陷入局部最优问题,综合分析了仿生智能聚类的特点,提出了一种混合交叉因子的蚁群聚类方法。该算法采用结合分阶段调整策略和启发式多点交叉策略的混合交叉因子,其中分阶段调整策略动态调整交叉点规模,显著降低交叉操作的无效性概率;启发式策略建立在适应度的基础上能有效地保留父代优秀基因。同时引用随机变异因子,进一步减少陷入局部优化的可能性。结合实例对算法进行了分析,结果表明了该算法在鲁棒性和聚类效果上都有所提高。
To deal with a problem that ant colony clustering easily trapped into local optimal, characteristics of the bionic intelligent clusterings are analyzed, and a hybrid crossover operator ofant colony clustering is proposed. Twocontrolmechanisms--thewell-phased control strategy, heuristic multipoint crossover strategy are built up a hybrid crossover operator, the well-phased control strategy dyna- mically adjust the crossover scale, which significantly reduced crossover operation of invalid probability, heuristic strategy is built on the basis of the fitness can be effectively retain father generation good genes. Furthermore, it combined with mutation operation avoiding local optimal. Analysis demonstrates that the proposed algorithm has improved the robust and the clustering results.
出处
《计算机工程与设计》
CSCD
北大核心
2011年第11期3840-3843,共4页
Computer Engineering and Design
基金
湖南省自然科学基金项目(07JJ6115)
智能制造湖南省高校重点实验室基金项目(2009IM06)
关键词
信息素
蚁群算法
聚类
混合交叉算子
变异算子
pheromone
ant colony algorithm
clustering
hybrid crossover operator
mutation operator