基于CA模型的转换坐标卡尔曼滤波  

Converted Measurement Kalman Filter Based on CA Model

作　　者：李一兵[1] 王博[1] 张兰[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院,哈尔滨150001

出　　处：《弹箭与制导学报》2008年第2期16-18,共3页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：国防基础科研基金资助

摘　　要：文中根据雷达对空间中的匀加速运动目标的跟踪要求,在三维空间中采用了转换坐标卡尔曼滤波算法(CMKFA),通过理论推导证明了当状态方程和测量方程满足一定条件时,转换坐标卡尔曼滤波算法是去偏转换测量值的线性无偏最小方差估计算法。在转换坐标卡尔曼滤波跟踪算法中,球坐标系下互相独立的观测噪声变换到直角坐标系下时将变成相关噪声,在解出此相关噪声的方差后即可按标准的卡尔曼滤波算法对目标进行跟踪,仿真结果显示,该算法收敛迅速,精度可以满足雷达对空间机动目标的跟踪要求。In this article, converted measurement Kalman filter （CMKF） was used in 3 dimensional space to meet the requirements of tracking a CV target, and it＇s proved that CMKFA is the least linear, unbiased variance algorithm based on the debiased converted measurements, provided some conditions are met. In CMKF algorithm, the independent observation noise in spheroidal coordinate system becomes correlated noise in rectangular coordinate system. Once the variance of the correlated observation noise has been solved.the standard Kalman filter can be used to track the target precisely, Simulation results indicate that the algorithm converges quickly and can track target with the required accuracy.

关 键 词：转换坐标卡尔曼 无偏最小方差估计 目标跟踪 

分 类 号：TN911[电子电信—通信与信息系统]