基于最大熵的机动目标状态融合估计  

Approach to State Fusion Estimation of Moving Targets Based on Maximum Entropy Principle

作　　者：刘唐兴[1] 孙裔申[1] 黄志良[1] 

机构地区：[1]中国电子科技集团公司第二十八研究所,南京210007

出　　处：《指挥信息系统与技术》2011年第2期19-22,70,共5页Command Information System and Technology

摘　　要：基于最大熵准则实现机动目标状态融合估计,针对聚类算法中隶属度计算与聚类中心耦合、类中心初始化不当的引入误差问题,提出向量解析方法解耦求解隶属度。仿真实验表明,该方法能避免聚类中心难选择问题,具有良好的估计精度。A new approach to the state fusion estimation of moving targets based on the maximum entropy principle is described. Targeting at the fact that the membership computation is coupled with the cluster centre in the cluster algorithm and the errors caused by improper cluster centre initialization, a vector analytic method is proposed to solve the membership. Simulation results indicate that the method avoids embarrassment： in cluster centre selection and has a good estimation accuracy.

关 键 词：数据融合 最大熵 模糊聚类 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]