一种目标跟踪滤波的新方法

曲洪权; 李少洪

doi:10.3724/SP.J.1146.2006.00738

一种目标跟踪滤波的新方法

doi: 10.3724/SP.J.1146.2006.00738

计量
- 文章访问数: 3007
- HTML全文浏览量: 77
- PDF下载量: 968
- 被引次数: 0
出版历程
- 收稿日期: 2006-05-29
- 修回日期: 2006-10-16
- 刊出日期: 2007-09-19

Novel Sequential Monte Carlo Method to Target Tracking

摘要

摘要: 对于目标跟踪系统，当观测不确定性相对系统不确定性较大时，如果采用EKF，UKF算法，由于概率密度函数(PDF)由高斯分布近似使真实的分布结构扭曲，导致系统性能下降或发散，采用粒子滤波时，因为系统不确定性相对观测不确定性较小，所以重采样会使粒子间的独立性消失，导致系统性能下降。为了提高目标跟踪的精度，该文给出一种SMCEKF及SMCUKF滤波算法，在SMC(Sequential Monte Carlo)算法中分别引入EKF及UKF，由独立滤波器更新和传播的随机采样点和相应权重来表示状态的PDF，由于初值和滤波都是独立的，所以确保了表示PDF的随机样值的独立性，在滤波器个数较少、计算量较小的情况下使滤波性能得到提高。文中给出了理论分析和仿真实例证明算法的有效性。
- 序列蒙特卡罗;随机采样;跟踪;UKF;粒子滤波
Abstract: EKF and UKF are often used in target tracking, but the required PDF is approximated by a Gaussian, which may be a gross distortion of the true underlying structure and may lead to filter divergence, especially in the situations where the uncertainty of the measurements is large compared to the uncertainty of process model of tracking. Resample introduces the problem of loss of diversity among the particles with particle filer because the uncertainty of process model is small compared to the uncertainty of the measurements. The SMCEKF and SMCUKF algorithms given in this paper ensure the independency of particles by introducing parallel independent EKF and UKF. The required density of the state vector is represented as a set of random samples and its weights, which is updated and propagated recursively on their own estimate. The performance of the filters is greatly superior to the standard EKF and UKF. Analysis and simulation results of the bearing only tracking problem have proved validity of the algorithms.

HTML全文

参考文献(1)

Gordon N J, Salmond D J and Smith A F M. Novel approach to nonlinear/non-Gaussian Bayesian state estimation [J]. IEE Proceedings.-F, 1993, 140(2): 107-113.[2]Julier S J and Uhlmann J K. Unscented filtering and nonlinear estimation [J].Proce. IEEE.2004, 92(3):401-422[3]Heine K. A survey of sequential Monte Carlo methods. Licentiate Thesis. Tampare University of Technology, 2005: 1-3, 53-55, 73-83.[4]Ristic B, Arulampalam S, and Gordon N. Beyond the Kalman Filter-Particle Filters for Tracking Applications [M]. Boston: Artech House, 2004: 32-45, 52-55.[5]Arulampalam M S, Maskell S, Gordon N, and Clapp T. A tutorial on particle filters for online nonlinear/non-Gaussian Bayessian tracking [J].IEEE Trans. on Signal Processing.2002, 50(2):174-188[6]Crisan D and Doucet A. A survey of convergence results on particle filtering methods for practitioners [J].IEEE Trans. on Signal Processing.2002, 50(3):736-746[7]Van der Merwe R, Doucet A, Nando de Freitas, and Eric Wan. The Unscented Particle Filter [R]. Technical Report CUED/F- INFENG/TR 380, Cambridge University Engineering Department, 2000: 7-10.[8]Nordlund P J and Gustafsson F. Sequential Monte Carlo filtering techniques applied to integrated navigation system [J]. Proceedings of the American Control Conference, Arlington, June, 2001: 4375-4380.[9]Nordlund P J. Recursive state estimation of nonlinear systems with applications to integrated navigation [R]. Technical Report LITH-ISY-R-2321, Department of Electronic Engineering. Linkoping University, SE-58183 Linkoping, Sweden, 2000: 42-52.[10]Ristic B and Arulampalam M S. Tracking a manoeuvring target using angle-only measurements: algorithms and performance [J].Signal Processing.2003, 83:1223-1238[11]Blackman S and Popoli R. Design and analysis of modern tracking systems [M]. Boston: Artech House, 1999: 178-181.

施引文献

资源附件(0)

访问统计