改进的种群分类蚁群算法及其应用  

Improved Character-Base Ant Colony Algorithm and Its Application

作　　者：刘芳[1] 李义杰[1] 

机构地区：[1]辽宁工程技术大学,计算机软件与理论辽宁葫芦岛125105

出　　处：《计算机系统应用》2010年第1期144-148,共5页Computer Systems & Applications

摘　　要：提出了一种改进的种群分类蚁群算法,该算法在种群分类的基础上,引入了蚂蚁的知觉感觉特性等。该算法能明显的防止蚁群算法可能出现早熟的问题,从而解决了传统蚁群算法加速收敛与早熟、停滞现象的矛盾。为了说明该算法的性能,将该算法应用到聚类分析算法中,设计了算法的模型以及算法步骤,并通过仿真实验证明了本算法的可行性和有效性。An improved Character-base Ant Colony Algorithm based on Character-base Ant Colony Algorithm and sentience consciousness characters is presented in this paper.It can significantly prevent precocity,then make a balance between accelerating convergence and averting precocity as well as stagnation.The algorithm is applied to clustering analysis and the algorithm steps of the model have been designed to illustrate the performance of the algorithm in this paper.The results of simulation on clustering analysis indicate that the new algorithm is feasible and effective.

关 键 词：蚁群算法 聚类分析 种群分类蚁群算法 蚂蚁的感觉知觉性 K均值算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] S682.29[自动化与计算机技术—计算机科学与技术]