基于极大似然的联合数据关联与空间配准  

Joint Data Association and Spatial Registration Based on ML

作　　者：周学平 李佳杰[1] 邹瑾涛 娄洋歌 ZHOU Xueping;LI Jiajie;ZOU Jintao;LOU Yangge(The 28th Research Institute of China Electronics Technology Group Corperation,Nanjing 210007 ,China)

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：《现代雷达》2019年第3期48-52,57,共6页Modern Radar

摘　　要：针对雷达量测存在漏探测情况下的雷达之间数据关联与雷达系统误差估计问题,提出了基于极大似然(ML)的联合数据关联与空间配准算法。该算法采用多维分配算法进行数据关联,使用ML对空间配准问题中的系统偏差进行估计,通过迭代进行数据关联和空间配准,最终得到收敛的系统偏差估计和关联正确率。仿真实验表明:在雷达探测存在漏探测时,文中提出的算法能够有效地进行数据关联和估计系统偏差,有较强的抗噪性和鲁棒性。Aiming at the data association and spatial registration under radar measurements with missing detection, a joint data association and spatial registration algorithm based on maximum likelihood(ML) algorithm is derived. Multidimensional assignment algorithm is adopted to associate measurement data, the radar system biases are estimated by ML algorithm, and the convergent system biases and associated matrix are obtained by jointly applying data associate and spatial registration iteratively. Simulation results show that under the environment that measurements may have missed detection, the proposed algorithm can estimate the system biases efficiently. It shows good robustness and anti-noise performance.

关 键 词：极大似然法 数据关联 空间配准 多维分配 

分 类 号：TN95[电子电信—信号与信息处理]