摘要
为了提高交互多模型算法的滤波精度,提出了基于无迹卡尔曼滤波(UKF)的交互多模型算法(IMM-UKF)。该算法融合了交互多模型算法对不同目标机动模式的自适应能力和UKF滤波精度高的优点。通过对机动目标跟踪的应用仿真,将该算法和基于扩展卡尔曼滤波(EKF)的交互多模型算法(IMM-EKF)进行了比较,仿真结果表明了IMM-UKF具有较好的跟踪性能,减小了机动目标跟踪的均方根误差。
An interacting multiple model algorithm based on the unscented Kalman （UKF） was proposed to improve the accuracy of interacting multiple model. The adaptive ability to various target maneuvering patterns was combined with the advantage of higher accuracy provided by UKE IMM-UKF was compared with interacting multiple model algorithm based on the extended Kalman （IMM-EKF） in maneuvering target tracking. Simulation results show that IMM-UKF is superior and its root mean square error （RMSE） is reduced.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第3期655-657,共3页
Journal of System Simulation
基金
国家自然科学基金资助项目(60501004)
航天科技创新基金资助项目(N7CH0003)
关键词
扩展卡尔曼滤波
无迹卡尔曼滤波
机动目标跟踪
交互多模型
extended kalman filter
unscented kalman filter
maneuvering target tracking
interacting multiple model
