基于合作目标的三维空间贝叶斯传感器配准算法

Registration Algorithms of Bayes Sensor Based on Cooperative Targets in ECEF

当存在合作目标时,基于合作目标的传感器配准算法往往优于基于多站测量的传感器配准算法。传统的基于合作目标的传感器配准算法将系统误差看作是一个确定未知量,对未知量的估计采用了非贝叶斯的参数估计算法,如极大似然法、最小二乘法等。当传感器量测噪声相对于传感器系统误差不可忽略时,上述算法估计效果较差。在此背景下,提出了基于合作目标的贝叶斯传感器配准算法,通过将传感器系统误差建模成未知的随机过程,并通过卡尔曼滤波消除量测噪声对系统误差估计带来的影响,实验结果表明,方法具有较好的实用性。

Sensorregistration methods based on cooperative targets performs better than those methods based on cooperative measurement.Traditional sensor registration methods based on cooperative targets such as ML and LS etc.regard system errors as unknown parameters,which are not effective when measurement noise exists.In order to solve this problem,Bayes sensor registration algorithm is proposed in this paper,which regards system error as random process and useKalman filter to weaken the influence of the noise.Finally,simulation results based on Monte Carlo show that Byes algorithm is more effective than traditional methods.