基于改进动态关联的多目标无源跟踪算法  

Passive Multitarget Tracking Algorithm Based on Improved Dynamic Association

作　　者：徐征[1] 曲长文[1] 李炳荣[1] 王昌海[1] 

机构地区：[1]海军航空工程学院电子信息工程系,山东烟台264001

出　　处：《雷达科学与技术》2013年第2期173-178,共6页Radar Science and Technology

基　　金：航空电子系统综合技术重点实验室和航空科学基金联合资助(No.20105584004)

摘　　要：针对状态-状态动态关联算法关联精度差,计算量大等缺点,提出了基于状态-量测动态关联的多目标只测角无源跟踪算法。该算法在航迹起始后,将目标状态估计直接与下一时刻由静态量测-量测数据关联确定的S元量测进行关联,利用二维分配算法得出关联对后,通过不敏卡尔曼滤波算法对关联出的方位角集合进行非线性滤波从而实现对与之关联目标的状态更新。仿真结果表明所提多目标跟踪算法能有效关联不同时刻源于同一目标的S元量测从而实现多目标跟踪且适合目标数目变化的情况,具有较好的工程应用前景。In order to solve the disadvantages of low association accuracy and high computation load for the state-state dynamic association, a bearings-only passive multitarget tracking algorithm is proposed, which is based on the state-measurement dynamic association. After track initialization, the proposed algorithm di- rectly associates the estimated states with the S-tuple of measurements conformed by the static measure- ment-measurement data association. Once the association pair is got by the two-dimensional assignment algo- rithm, the unscented Kalman filter（UKF） is used to update the state of the target using the corresponding bearings set associated with the target. Simulation results indicate that the proposed multitarget tracking al gorithm can effectively associate the S-tuple of measurements originated from the same target at different times to achieve multitarget tracking and is suitable for the circumstances of varying target numbers, which indicates its good prospect for the practical applications.

关 键 词：无源跟踪 多目标 只测角 动态关联 不敏卡尔曼滤波 

分 类 号：TN958.97[电子电信—信号与信息处理]