基于时间序列的自适应加权被动定位

Adaptive Weighted Passive Location Based on Time Series

现代海战场电磁环境日益复杂 ,利用卡尔曼滤波进行纯方位跟踪定位具有很强的隐蔽性 ,但多传感器固定权值融合算法在复杂的海战环境下有一定的缺陷。该文针对纯方位定位数据融合特点 ,运用时间序列估计传感器方差实现传感器权值的在线调整 ,仿真结果表明该算法较固定权值跟踪精度有明显提高 ,对提高纯方位目标定位的实用性及适应性有一定的意义。

Modern sea battle's electromagnetic surroundings become more and more complicated.Though using Kalman filter to estimate location and track with bearing only measurements is very conceal, but fixed weight fusion algorithm has some disadvantages in complicated sea battle field. According to the characteristics of data fusion based bearing only measurements, this paper uses time series to estimate sensors' variances and adjust their weight. The simulation results show that in contrast to fixed weight algorithm, target-tracking accuracy can be greatly improved.It has some meaning for improving passive location's practicality and adaptability.