基于UKF算法的被动目标跟踪

对于单站的被动目标跟踪,在笛卡儿坐标系下建立跟踪模型,并用扩展的卡尔曼滤波(EKF)进行预测,得到的结果通常是不稳定且容易发散的。针对这种情况,提出了在修正的极坐标系下建立状态模型,摒弃传统的EKF算法,采用无迹卡尔曼滤波(UKF)算法,通过采样逼近非线性函数。数字仿真结果表明:在修正的极坐标中利用UKF算法得到的结果比EKF算法具有更快的收敛速度和更高的估计精度,且稳定性更好。

Bearings-only Tracking Based on Unscented Kalman Filter

In single station passive tracking,when using Cartesian coordinates for establishing tracking models and the Extended Kalman Filter(EKF) for predicting,the results obtained are often unstable and easily get divergence.In view of this situation,we established the state model in the modified polar coordinates,and used Unscented Kalman Filter(UKF) in place of EKF.This algorithm approximates the nonlinear function through sampling.The simulation results proved that using the UKF algorithm in the modified polar ...