摘要
提出了一种利用基于距离参数的最小二乘滤波器群的纯方位跟踪目标参数估计器 ,该估计器是由若干个处地不同的距离区间上的加权最小二乘滤波器组成 .通过假设预测方位和实测方位差值服从零均值的高斯分布 ,利用贝叶斯理论来修正各滤波器的权重 .每获得一个量测方位值 ,各个滤波器的权重将修正一次 ,且各个滤波器的权重之和应为 1,当其权重小于某一阈值时 ,则该滤波器停止工作 ,最终剩余的滤波器将给出准确的目标参数的估计值 .计算机模拟计算表明该方法在收敛速度。
In this paper, a new bearings only tracking approach, which consists of a set of weighted least squares filters(LSF) each with a different initial range estimate, is proposed. And the weight of each filter is updated using Bayes theory based on the assumption that the difference between estimate and measurement bearings obeys Gaussian distributions with zero mean error. A lot of simulation experiments demonstrate that the new approach overcomes the problems exhibited with the existing Extended Kalman Tracker and the traditional least squared tracker. Its convergence and stability have been improved significantly.
出处
《弹道学报》
CSCD
北大核心
2003年第1期5-10,16,共7页
Journal of Ballistics
基金
国家自然科学基金资助项目 ( 6 95 72 0 15 )