单站目标被动跟踪的滤波器方法研究

Filters Analysis for Passive Tracking by One Sensor

为了实现对单站目标的被动跟踪,分析并比较了扩展Kalman滤波器和粒子滤波器在非线性估计方面的性能,并且针对粒子滤波器存在的粒子退化现象,引入改进的重采样算法和基于无迹变换的滤波算法.仿真实验分别比较了几种滤波器在目标做匀速、匀加速、变加速情况下距离和速度滤波的均方根误差,结果表明粒子滤波器滤波性能优于扩展的Kalman滤波器,改进的重采样算法和基于无迹变换的粒子滤波器可以有效改善估计精度.

For targets passive tracking by only one sensor,performance of Extend Kalman Filter and Particle Filter on nonlinear estimation is compared after deep analysis.To eliminate degeneration existing in Particle Filter,an improved algorithm of re-sampling and Unscented Transformation are adopted.Through emulation,targets' length and speed root-mean-square error estimated by the filters above are compared on the conditions of uniform motion,uniform acceleration,and fluctuating acceleration respectively.The result...