基于平方根迭代UPF滤波的无源定位算法研究

无源定位是基本的导航问题之一,是机动目标跟踪问题研究的基础。针对固定辐射源的无源定位问题,常见的标准粒子滤波器存在收敛速度慢的缺陷,设计了一种基于平方根迭代的UPF滤波器,采用平方根迭代UKF算法作为提议函数,改进了传统PF算法,给出了一种SRUPF算法的设计方案。以一个典型的二维平面无源定位问题对算法进行仿真,仿真结果表明平方根迭代UPF滤波器和标准粒子滤波器相比,具有更高的估计精度,同时较一般UPF滤波器具有更好的数值鲁棒性和更快的收敛速度。

Passive Position Localization Algorithm Using Square Root Iterative UPF Filter

Position localization is one of the most basic navigation problems and the foundation of maneuvering target tracking problems. The generic particle filter converges slowly in the passive position localization problem of stationary emitter. Aiming at the defect, a square-root iterative unscented particle filter (UPF) is introduced in the paper, which uses the square root iterative UKF algorithm as the proposed function to improve the traditional PF algorithm and gives a design scheme of the SRUPF algorithm. A typical two-dimensional passive location problem is used to simulate the proposed algorithm. Compared with standard UPF filter, the proposed method has the advantage of higher accuracy, faster convergence speed and better numerical robustness.