摘要
针对磁性椭球体的跟踪问题具有高度非线性的特点,将磁性椭球体跟踪问题归结为动态系统的贝叶斯估计问题,提出了利用磁场梯度张量信息的递归方法来估计其运动轨迹和磁矩参数;并在此基础上应用高斯混合采样粒子滤波算法实现了目标跟踪.使用模拟磁性椭球体进行仿真实验,结果表明:提出的跟踪方法解决了磁性椭球体跟踪问题,且高斯混合采样粒子滤波算法相比其他算法有更好的性能和更小的计算量.
Magnetic ellipsoid tracking problem was characterized by high nonlinearity. The magnetic ellipsoid tracking was formulated as Bayesian estimation for dynamic system. A recursive approach was proposed to estimate the trajectory and the magnetic moment component of the target using the data collected with a magnetic gradiometer tensor. Based on the proposal method, the Gaussian mix- ture sigma point particle filter (GMSPPF) algorithm was adopted to realize target tracking. The per- formance of the method was evaluated through the simulation experiment. The results indicate that the method can achieve the magnetic ellipsoid tracking, and the performance of GMSPPF performs is better in estimation and computational efficiency than that of other algorithms.
出处
《华中科技大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2013年第11期103-107,共5页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家自然科学基金资助项目(51109215)
关键词
磁场梯度张量
非线性
跟踪
贝叶斯估计
高斯混合采样粒子滤波
椭球体
magnetic gradiometer tensor
nonlinearity
tracking
Bayesian estimation
Gaussian mix-ture sigma point particle filter (GMSPPF) ~ ellipsoid