基于高斯粒子滤波的机载GMTI雷达跟踪

Airborne Radar for GMTI Radar Tracking Filtering Based on Gaussian Particle Filter

在使用机载雷达对地面目标进行跟踪过程中,常规粒子滤波算法时间复杂度高,误差性能改善幅度不明显。在状态向量服从高斯分布的条件下,将高斯粒子滤波算法应用到机载GMTI雷达非线性滤波过程中。该方法可以根据状态向量的二阶统计量抽取粒子,粒子权值不需要递推计算。相对于常规粒子滤波算法,降低了时间复杂度,并且能够在一定程度上提高误差性能。仿真结果证明了该方法的有效性。

The normal particle filter has high time complexity and low improvement in accuracy when it is used in the airborne radar for Ground Moving Target Indicator（GMTI） tracking.Gaussian particle filter was introduced to the airborne radar for GMTI tracking as the state vector subjected to Gaussian distribution.The Gaussian particle filter could extract particles according to the second order statistics directly,and the weights of particles need not to be evaluated recursively.Compared with the normal particle filter,the method can decrease the time complexity and improve accuracy to some extent.Simulation results show that the new method is effective.