混合线性/非线性模型的准高斯Rao-Blackwellized粒子滤波法

 针对混合线性/非线性模型，提出一种新的递推估计滤波算法，称为准高斯Rao-Blackwellized粒子滤波器（Q-GRBPF）。算法采用Rao-Blackwellized思想，将线性状态与非线性状态进行分离，对非线性状态运用准高斯粒子滤波（Q-GPF）算法进行估计，并将其后验分布近似为单个高斯分布，再利用非线性状态的估计值对线性状态进行卡尔曼滤波（KF）估计。将Q-GRBPF应用于目标跟踪的仿真结果表明，与Rao-Blackwellized粒子滤波器（RBPF）相比，Q-GRBPF在保证估计精度的前提下有效降低了计算复杂度，计算时间约为RBPF的58%；与Q-GPF相比，x坐标与y坐标的估计精度分别提升了45％和30％，而计算时间也节省了约30％。

Quasi-Gaussian Rao-Blackwellized Particle Filter for Mixed Linear/Nonlinear State Space Models

A new recursive estimation algorithm,called the quasi-Gaussian Rao-Blackwellized particle filter(Q-GRBPF),is proposed for filtering mixed linear/nonlinear state space models.The algorithm utilizes the idea of Rao-Blackwellized to separate the linear and nonlinear states.For the nonlinear states,the posterior distributions of the estimates,which are achieved by the quasi-Gaussian particle filter(Q-GPF),are approximated as Gaussian distributions.Also,the linear states are estimated by the Kalman filter(KF) with the estimated nonlinear states.The simulation results of the proposed method applying to target tracking show that the proposed method only consumes 58% of the computing time required by the RBPF.Furthermore,compared with Q-GPF,the tracking accuracies of the proposed method for estimating x and y coordinate locations of the tracked target are respectively increased by 45% and 30% while 30% computing time is saved.