摘要
基于末制导雷达搜索舰艇编队目标时获得的目标大角域高分辨距离像(HRRP)信息,该文建立了描述单个HRRP样本内向量之间统计关系的"各态历经"空间隐马尔可夫模型(SHMM)和描述HRRP样本之间统计关系的"从左到右"时间隐马尔可夫模型(THMM)。与对一类目标全方位角训练数据只建立一个THMM模型的方法相比,该方法充分利用目标的大角域HRRP信息,提高了识别性能。通过对5类舰船目标的仿真和3类民用船只的外场实测数据分析表明该方法的有效性。
Based on the target large angle domain High Resolution Range Profile（HRRP） information of the ship formation obtained by the terminal guidance radar during its search phase, this study establishes an ＂ergodic＂ Spatial Hidden Markov Model（SHMM） which describes statistical relationship between the vectors in a single HRRP sample and a ＂left to right＂ Temporal HMM（THMM） which describes statistical relationship between HRRP samples. In comparison with the method that it only establishes a THMM model with the training data of all-round angle of one target, the proposed method makes full use of the target HRRP information of large angle domain and can improve the recognition performance. Through the simulation of the five types of ship target and the field measured data analysis of three kinds of civilian vessels show that the effectiveness of the proposed method.
出处
《电子与信息学报》
EI
CSCD
北大核心
2015年第4期926-932,共7页
Journal of Electronics & Information Technology
基金
上海市科学技术委员会资助课题(11DZ2260800)
省部级基金资助课题
关键词
雷达高分辨距离像
空间隐马尔可夫模型
时间隐马尔可夫模型
编队目标识别
Radar High Resolution Range Profile（HRRP）
Spatial Hidden Markov Model（SHMM）
Temporal Hidden Markov Model（THMM）
Formation target recognition
