一种改进粒子滤波器在雷达目标跟踪中的应用  

Unscented Particle Filter and Its Application to Radar Target Tracking

作　　者：吕学斌[1] 周群彪[1] 陈正茂[1] 赵明华[1] 

机构地区：[1]四川大学图形图象研究所,成都610064

出　　处：《系统仿真学报》2007年第9期2097-2100,共4页Journal of System Simulation

基　　金：国家级火炬计划项目(2001EB001119)

摘　　要：在实际雷达目标跟踪系统中,雷达量测常受到闪烁噪声干扰,传统卡尔曼或扩展的卡尔曼滤波算法在闪烁噪声环境下,滤波性能将急剧下降甚至滤波发散。提出了将粒子滤波与无迹变换结合的改进粒子滤波算法UPF(uncented particle filter)应用在雷达目标跟踪中,解决了闪烁噪声情况下雷达目标跟踪问题。仿真结果表明,在高斯条件下扩展的卡尔曼算法和基于无迹变换的粒子滤波算法跟踪性能相近,但在闪烁噪声环境下,随着闪烁影响的增强,扩展的卡尔曼算法跟踪性能严重下降,而UPF算法能保持较好的跟踪精度。In real radar target tracking system, the measure data of radar are often disturbed by the glint noise. If still using the kalman or extended kalman filter, a large error will be product. An algorithm based on the unscented particle filter was proposed which was introduced to radar target tracking based on the glinst noise statistical model. Simulation result shows that in the Gaussian environment both extended kalman filter and unscented particle filter have almost the same tracking accuracy, and that in glint noise environment unscented particle filter has also good accuracy, while the extended kalman filter＇s performance degrades severely as the glint effect increasing.

关 键 词：目标跟踪 粒子滤波 扩展卡尔曼滤波 UKF UPF 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]