多传感器粒子滤波融合跟踪算法

Multi-sensor Fusion Tracking Algorithm Based on Particle Filter

对于非线性非高斯环境中的多传感器分布式状态估计问题,提出了一种基于二阶中心差分粒子滤波方法的融合跟踪算法。通过对量测方程的非线性分析,利用粒子滤波器计算目标状态估计值,以在线自适应加权融合算法的方式得到系统最优估计。仿真结果表明,与采用扩展卡尔曼滤波的方法相比,该算法具有更高的估计精度。

In order to solve the distributed multi-sensor state estimation problem of non-Gaussian nonlinear system,a fusion tracking algorithm based on second-order central difference particle filter is proposed. It uses particle filter to calculate state estimated values by the nonlinear analysis of measurement equation,and then the system optimal estimation is obtained in the adaptive weighted fusion algorithm way. The simulation results show that compared with the extended Kalman filter,the proposed algorithm improves the estimation accuracy.