三维纯角度被动跟踪定位的最小二乘-卡尔曼滤波算法

LS-Kalman algorithm for passive target location and tracking with bearing-only measurements

利用角度信息估计出目标的距离和速度实质上是一个非线性状态估计问题 ,经典的扩展卡尔曼滤波算法性能很不稳定。文中首先根据静态估计理论推导出了在某一时刻目标位置的最小二乘解 ,然后将其作为卡尔曼滤波的测量值进行滤波 ,作进一步的数据处理 ,以提高估计精度。为了避免测量误差的相关性 ,分别在x ,y ,z方向上进行滤波 ,简化了算法 ,提高系统的定位精度。仿真结果表明这一算法是简单而有效的。

Bearing-only location is a nonlinear state estimation in essence. Classical extended Kalman filter is erratic. In this paper, expressions for the least squares estimation of target location and its variances have been derived by the static estimation theory. Then Kalman filter is used in x, y, z directions respectively to improve the location accuracy. Simulation shows that it is simple and efficient.