摘要
针对磁梯度张量直接反演跟踪方法对噪声十分敏感的问题,提出了基于正则化粒子滤波(RPF)的磁梯度张量跟踪方法。该方法建立了包括目标位置、速度和磁矩等状态变量的状态方程,给出了表征状态向量与磁梯度张量、磁场三分量测量值关系的量测方程,加入了重采样等克服粒子退化现象的方法,并通过RPF方法克服粒子衰竭问题。通过仿真实验证明了RPF跟踪方法的精度远远高于直接反演跟踪方法,具有更强的抗噪声干扰能力,可以有效完成磁梯度张量跟踪。
Aiming at the problem that direct inversion tracking method of magnetic gradient tensor was too sen-sitive to noise,tracking method of magnetic gradient tensor based on RPF was put forward in this paper.The state equation was built,which included the state variables of position,velocity and magnetic moment of obj ect. The measurement equation represented the relation between state variables and measurement values of magnetic gradient tensor and three magnetic components.Resampling was used to prevent particle degradation,and the problem of particle lack was overcome by RPF.It was proved that the accuracy of the tracking method based on RPF was much higher than that of direct inversion tracking method by simulation experiment.
出处
《探测与控制学报》
CSCD
北大核心
2014年第2期50-52,58,共4页
Journal of Detection & Control
关键词
正则化粒子滤波
跟踪方法
磁梯度张量
重采样方法
rgularized particle
filter（RPF）
tracking method
magnetic gradient tensor
resampling method
