基于PF-Singer的机动目标跟踪算法  

Maneuvering Target Tracking Algorithm Based on PF-Singer

作　　者：张楠[1] 徐毓[1] 张萍[1] 彭焱[1] 

机构地区：[1]空军雷达学院信息与指挥自动化系,湖北武汉430019

出　　处：《计算机测量与控制》2006年第2期242-243,255,共3页Computer Measurement &Control

基　　金：国防预研基金项目(51421040103JB4902)

摘　　要：Singer模型是典型的全局统计模型,其严重缺陷在于所采用的零均值时间相关模型和标准卡尔曼滤波算法不能完成对机动目标状态的正确估计;只有当目标做匀速直线运动时,动态误差的稳态值才为零,否则不为零;采用PF-Singer算法对机动目标进行跟踪,能够有效解决传统Singer模型存在的问题,提高其跟踪精度;通过仿真试验证实了该算法的有效性。Singer model is a classic whole statistic model, it is fatal limitation exits in the use of time-correlated model with zero-mean and Kalman filter that can＇t effectively solve maneuvering target tracking problem. Merely when the target bears a constant velocity, the steady state value of dynamic error presents zero, otherwise reverse. Tracking maneuvering target using PF-Singer algorithm can resolve the limitation of Singer model and advance accuracy of tracking. The simulation shows its effectiveness.

关 键 词：机动目标 目标跟踪 PF—Singer算法 

分 类 号：TN953[电子电信—信号与信息处理]