基于直觉模糊核聚类的弹道中段目标识别方法  

Technique for target recognition in ballistic midcourse based on intuitionistic fuzzy kernel clustering

作　　者：范成礼[1] 邢清华[1] 付强[1] 范学渊[1] 

机构地区：[1]空军工程大学防空反导学院,陕西西安710051

出　　处：《系统工程与电子技术》2013年第7期1362-1367,共6页Systems Engineering and Electronics

基　　金：国家自然科学基金(61272011)资助课题

摘　　要：针对现有的模糊核聚类算法性能的问题,汲取直觉模糊c-均值聚类(intuitionistic fuzzy c-means,IFCM)算法的动态聚类特性优势,引入高斯核函数,改良归一化条件,提出直觉模糊核c-均值聚类(intuitionisticfuzzy kernel c-means,IFKCM)算法,并通过实际数据测试,证实了该算法的可行性和有效性。最后,根据弹道中段目标识别仿真系统的要求及弹道目标识别的特点,设计并实现了基于直觉模糊核c-均值聚类的弹道中段目标识别(intuitionistic fuzzy kernel c-means-target recognition in ballistic midcourse,IFKCM-TRBM)原型系统,仿真实验及对比分析充分表明该原型系统的稳健可行性,为弹道中段目标识别提出了一种新的参考和尝试。A kernel based intuitionistic fuzzy clustering algorithm named IFKCM is proposed on the basis of analyzing the deficiency of the existing clustering algorithm.Gauss kernel is introduced,the constraint condition is improved and the property superiority of dynamic clustering performance is used,which is also the superiority in intuitionistic fuzzy c-means（IFCM） algorithm.Then the experimental result proves its effectiveness.Subsequently,according to the requirement of target recognition in ballistic midcourse simulation system and the character of the ballistic target recognition,the simulation system named intuitionistic fuzzy kernel c-means-target recognition in ballistic midcourse（IFKCM-TRBM） is designed and realized.The results of simulation show that the system is reliable and can support the research of target recognition in ballistic midcourse.

关 键 词：直觉模糊聚类 模糊核c-均值 高斯核函数 弹道中段 目标识别 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]