基于量化新息的容积粒子滤波融合目标跟踪算法

针对现有非线性网络化目标跟踪融合算法存在的精度低和实用性差等不足,以一类带有噪声相关的非线性网络化目标跟踪系统为对象,研究基于测量新息量化策略和容积粒子滤波（Cubature particle filter,CPF）的目标跟踪融合算法.首先,利用状态方程恒等变换和矩阵相似变换理论解除过程噪声与测量噪声以及测量噪声之间的相关性;其次,各个传感器节点采用自适应策略量化局部测量新息并将其发送到融合中心（Fusion center,FC）;随后,在集中式融合框架下采用容积粒子滤波器设计基于测量值扩维的量化融合跟踪算法,进而给出相应的顺序滤波量化融合算法,上述算法可有效解决因自适应量化引起的非高斯问题;...

Full description

Saved in:
Bibliographic Details
Published in自动化学报 Vol. 40; no. 9; pp. 1867 - 1874
Main Author 徐小良 汤显峰 葛泉波 管冰蕾
Format Journal Article
LanguageChinese
Published 杭州电子科技大学计算机学院 杭州 310018%浙江大学教育学院杭州310028%杭州电子科技大学自动化学院杭州310018%宁波工程学院电子与信息工程学院 宁波 315016 2014
Subjects
噪声相关
容积粒子滤波
无线传感器网络
比特位量化
目标跟踪
cubature particle filters (CPF)
Wireless sensor networks (WSN)
噪声相关
target tracking
无线传感器网络
容积粒子滤波
目标跟踪
比特位量化
bits quantization
correlated noises
Online AccessGet full text
ISSN0254-4156
1874-1029
DOI10.3724/SP.J.1004.2014.01867

Cover

More Information
Summary:针对现有非线性网络化目标跟踪融合算法存在的精度低和实用性差等不足,以一类带有噪声相关的非线性网络化目标跟踪系统为对象,研究基于测量新息量化策略和容积粒子滤波（Cubature particle filter,CPF）的目标跟踪融合算法.首先,利用状态方程恒等变换和矩阵相似变换理论解除过程噪声与测量噪声以及测量噪声之间的相关性;其次,各个传感器节点采用自适应策略量化局部测量新息并将其发送到融合中心（Fusion center,FC）;随后,在集中式融合框架下采用容积粒子滤波器设计基于测量值扩维的量化融合跟踪算法,进而给出相应的顺序滤波量化融合算法,上述算法可有效解决因自适应量化引起的非高斯问题;最后,通过两个计算机仿真实验验证了所提出跟踪算法的有效性.
Bibliography:For the nonlinear networked target tracking system with arbitrarily correlated noises, target tracking fusion algorithms based on quantized innovation and cubature particle filter （CPF） are researched in order to overcome the shortages of the existing methods, which have low precision and poor real-time performance. Firstly, identical transformation of state equation and matrix similarity transformation theory are used for arbitrary noises decorrelation. And then, each sensor node adopts an adaptive quantization strategy to quantize its innovations, and transmits them to the fusion center （FC）. Subsequently, the nom-Gaussian problems caused by quantization are solved by using CPF to design a fusion tracking algorithm with augmented measurements in the centralized fusion framework. Moreover, its corresponding sequential fusion form is developed. Finally, two computer simulation experiments show the effectiveness of the proposed method.
XU Xiao-Liang TANG Xian-Feng GE Quan-Bo GUAN Bing-Lei（ 1. School of Compute
ISSN:0254-4156
1874-1029
DOI:10.3724/SP.J.1004.2014.01867