基于Unscented卡尔曼滤波的目标跟踪仿真分析  

Simulation Analysis of Target Tracking Method Based on Unscented Kalman Filter

作　　者：时少旺[1] 谭守林[1] 

机构地区：[1]第二炮兵工程学院作战保障系,陕西西安710025

出　　处：《郑州大学学报（工学版）》2008年第3期47-50,共4页Journal of Zhengzhou University（Engineering Science）

摘　　要：卡尔曼滤波技术是目标跟踪的理论基础之一.在工程应用中,多采用非线性滤波的方法对目标进行跟踪.介绍了Unscented变换,通过使用Unscented卡尔曼滤波(UKF)对目标跟踪进行仿真实验,Un-scented卡尔曼滤波有效地克服了传统的扩展卡尔曼滤波算法(EKF)的缺陷,给出了最佳估计的二阶近似,提高了对目标跟踪的精度和稳定性.仿真结果表明,该算法对目标在位置、速度跟踪方面均有良好的效果;针对目标机动,采用多模型滤波算法能够实现目标机动的精确跟踪.Kalman Filter technology is one of the theory bases of target tracking. In engineering application, nonlinear filter algorithm is mostly used for target tracking. In this paper, Unscented Kalman Filter（UKF） is introduced to make simulation analysis of target tracking. For target tracking, The UKF overcomes the disadvantages of conventional Extended Kalman Filter （EKF） effectively and gives the optimal estimation approximation. The UKF improves the accuracy and stability. The simulation shows that the model proposed in this paper can well estimate the position and velocity of target tracking. For target manoeuver, accurate tracking is obtained with the multiple model algorithm.

关 键 词：无味变换 目标跟踪 UNSCENTED卡尔曼滤波 多模型算法 

分 类 号：TN957.51[电子电信—信号与信息处理]