基于自适应和及时繁殖策略菌群优化聚类算法  

Adaptive and timely multiplying BFO clustering algorithm

作　　者：赵明茹[1,2] 唐恒亮[1,2] 郭键[2] 孙媛[2] 

机构地区：[1]北京工业大学计算机学院多媒体与智能软件技术北京市重点实验室,北京100124 [2]北京物资学院北京市智能物流系统重点实验室,北京101149

出　　处：《计算机应用研究》2014年第8期2283-2286,共4页Application Research of Computers

基　　金：北京物资学院青年科研基金资助项目(2012XJQN016);北京市哲学和社会科学规划资助项目(13JGC099);中国博士后科学基金资助项目(2013M530498)

摘　　要：针对菌群优化算法求解优化问题时收敛速度较慢、易陷入局部最优值的缺点,提出一种基于自适应和及时繁殖策略的菌群优化聚类算法。在算法中根据迁徙的次数改变步长,适应度最大的细菌马上进行繁殖。为了进行对比实验,在二维平面随机生成了100个点,将这100个点聚成两类和三类。实验结果证明自适应和及时繁殖这两种策略都可以明显提高菌群优化算法的收敛速度。In order to slove the problem of slow convergent speed and just getting local optimum solution easily in solving optimization problems,this paper proposed a modified bacterial foraging optimization clustering algorithm with the adaptive and timely reproductive strategies. First,the algorithm changed the step size according to the times of migration. Then,the bacteria of largest fitness bred immediately. For the purpose of carrying out comparative experiments,the paper generated 100 two-dimensional points randomly and clustered them into two clusters and three clusters. Experimental results show that the adaptive and timely reproductive strategies can significantly improve the convergence speed of clustering algorithm.

关 键 词：聚类 菌群优化算法 自适应策略 及时繁殖策略 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] TP301.6[自动化与计算机技术—计算机科学与技术]