闪烁噪声环境下机动目标跟踪的改进的高斯-厄米特粒子滤波

An IGHPF algorithm of maneuvering target tracking in glint environment

针对闪烁噪声环境下机动目标跟踪的非线性、非高斯问题,提出了一种改进的高斯-厄米特粒子滤波算法。和传统的高斯-厄米特粒子滤波算法相比,在生成粒子集时,改进的高斯-厄米特粒子滤波算法采用高斯-厄米特滤波对当前时刻的各个粒子进行估计,将得到的估计值和协方差直接作为粒子滤波算法的粒子集及相应的协方差。仿真结果表明,改进的高斯-厄米特粒子滤波算法对闪烁噪声环境下的机动目标能够进行有效的跟踪,提高了跟踪精度。

Abstract： An improved Gaussian Hermitian particle filter （IGHPF） is proposed in order to solve ma- neuvering target tracking problem of nonlinear non-Gaussian under glint noise. Compared with tradition- al Gaussian Hermitian particle filter （GHPF）, when producing particle set, IGHPF directly uses Gauss- ian Hermitian Filter （GHF） to estimate the current particles and the obtained estimation value and co- variance are as the particle set and covariance of particle filer algorithms. Simulation results show that IGHPF can effectively track maneuvering target under glint noise, improving tracking precision.