城市战场目标分群的组合聚类方法  

Combination Clustering Method for Target Grouping in Urban Battlefield

作　　者：张松良[1,2] 王付明[1] 鲁柯 谢永生[1] 

机构地区：[1]解放军信息工程大学 [2]66322部队 [3]河南陆军预备役136师

出　　处：《指挥控制与仿真》2009年第5期37-41,共5页Command Control & Simulation

摘　　要：为了消除或减少城市作战中灰色情报信息对态势评估中目标分群的影响,提出一种基于Chameleon算法的组合聚类方法。根据目标属性间的区间灰数关联度和目标间的距离系数得到相似度矩阵,利用Chameleon算法对目标进行初始分群,然后,对分群中各目标的属性历史数据和分群中心的属性历史数据分别进行区间灰数向量关联分析,实现了目标分群的动态维护。通过算例验证了方法的有效性。In order to decrease the influence of gray information which is used for target grouping about situation assessment in urban battlefield, a combination clustering method based on chameleon algorithm is presented. The method combines the inter-zone gray correlation degree of target and distance coefficient of target to build a similar degree matrix, and uses the chameleon algorithm to realize the initial target grouping. Through the inter-zone gray vector relational analysis on target＇s attribute historical data and the center of target group＇s attribute historical data, the initial target grouping＇s maintenance is realized dynamically. The numerical example result shows the effectiveness.

关 键 词：灰数 灰色关联度 CHAMELEON算法 目标分群 组合聚类 

分 类 号：E911[军事] N945.11[自然科学总论—系统科学]