基于耗散结构理论的极大熵目标分群算法  

Maximum entropy object grouping algorithm based on dissipative structure theory

作　　者：李赟[1] 老松杨[1] 刘钢[1] 王石 

机构地区：[1]国防科学技术大学信息系统与管理学院,长沙410073 [2]总装备部武汉军代局长沙军代室,长沙410014

出　　处：《系统工程理论与实践》2012年第12期2816-2824,共9页Systems Engineering-Theory & Practice

摘　　要：为快速准确实施战场目标分群,从信息交互角度,结合耗散结构理论将战场态势认为是能够出现"有序"状态的动态开放系统.据此,建立用于战场目标分群的OG-Brusselator模型,并且改进了MEC算法,在航路勾径基础上提出航路股径概念统一目标彼此间距离和运动方向两要素的量纲,将其和目标与分群中心隶属度分别作为影响目标群属关系明晰度的可量化正、负熵指标,提出以OG-Brusselator模型控制算法迭代的目标分群算法DS-MEOC.最后针对空中目标进行分群实验分析,结果表明DS-MEOC算法有效可行,相比MEC算法,能够提供更合理的目标分群方案.To group battle object rapidly and exactly, consider battle situation as a dynamic open system in terms of information exchange and with the dissipative structure theory, in which ＂ordered＂ status could appear. On these grounds, established the OG-Brusselator mode for battle object grouping and improved the algorithm MEC, then proposed the concept of fairway lengthways-cut based on the fairway crosscut to unify dimensions of two elements： the distance between objects and their direction of movement, and set them together with the memberships of object to grouping center as the measurable positive and negative entropy indicators respectively that affecting the clarity of relationship between object and group, then proposed object grouping algorithm DS-MEOC using OG-Brusselator to control the iteration. The experiment and analysis of air targets grouping proved that DS-MEOC is effective, feasible and could provide more reasonable object grouping program as compared with MEC.

关 键 词：战场态势 目标分群 耗散结构 极大熵 

分 类 号：E919[军事]