【作者】 罗四； 李少洪；

【Author】 Luo Si Li Shaohong (School of Electronics and Information Engineering,Beijing Univ.of Aeronautics and Astronautics,Beijing 100083,China)

【机构】 北京航空航天大学电子信息工程学院；

【摘要】 针对雷达目标的高分率距离像的特点,对基于灰色关联分析方法的自动目标识别进行了深入研究。结合灰色关联分析法,选取雷达目标一维距离像作为参考数列和比较数列,并对序列进行无量纲化和归一化预处理,给出了灰色关联差异信息矩阵、灰色关联系数和灰色关联度的建立方法,对灰色关联系数的选取作了详尽分析,并且基于最大关联准则,依据灰色关联序列来进行分类判别,识别了四类不同飞机目标的高分辨率距离像。同时,本文也提出了一种改进的灰色关联分析法用于雷达目标识别,作了对比分析,指出了进一步研究的方向。实验结果表明,灰色关联分析法用于基于雷达高分辨率距离像的自动目标识别有较高的识别率,有一定的稳健性,有较好的可行性和有效性。更多还原

【Abstract】 Automatic target recognition(ATR) based on grey relational analysis(GRA) is deeply studied in accordance with the characteristics of high resolution range profiles(HRRPs) of radar targets.One dimensional range profiles of radar targets are chosen as a reference sequence and comparative sequences according as grey relational analysis.Data sequences are preprocessed by dimensionless standardization and normalization firstly,and then the establishment of grey relational difference information matrix,grey relational coefficient(GRC) and grey relational grade(GRG) is presented concretely. Selection of grey relational distinguishing coefficient is discussed in great detail.After all,high resolution range profiles of four different aircrafts are classified and identified according to grey relational rank ordinal based on the criterion of maximum relation.Meanwhile,a modified approach of grey relational analysis is proposed to recognize the radar targets and is analyzed comparatively.The aspects of future research are pointed as well.Experimental results demonstrate that GRA used in ATR based on HRRPs can yield good performance of high recognition ratio and it is robust to some extent.The simulation results also show the validity and feasibility of the novel method.更多还原