特征敏感度与模糊度的弹道目标聚类识别评估

Clustering Identification Evaluation of Ballistic Missile Target Based on Sensitivity and Fuzzy Degree

针对现有弹道导弹目标聚类识别算法中,很少同时考虑目标特征敏感度与模糊度的问题,提出了一种基于特征敏感度与模糊度的弹道目标聚类识别算法。该算法通过目标特征类内类间距离,定义目标特征聚类识别敏感度,将两类目标形成的类看成两个圆,利用圆之间的位置关系,定义目标特征聚类识别模糊度,将所定义的特征敏感度与特征模糊度合理组合,作为目标特征对聚类识别好坏的评估因子。仿真实验从特征评估分析和特征组合分析两个方面验证了算法的有效性。

In existing clustering identification algorithms of ballistic missile target,it is always seldom considered the sensitivity and fuzzy degree of target,this paper proposes a new algorithm. This algorithm first defines the sensitivity of clustering identification by within-class and between-clas of target feature. And then the two kinds of target as two roundity are considered,and the fuzzy degree of clustering identification sensitivity is defined by using position relationship of two roundity. Finally,the evaluation factor is defined by reasonable combining sensitivity and fuzzy degree. Simulation results show the effectiveness of the proposed algorithm.