Generalised particle filters with Gaussian mixtures

被引:12
作者
Crisan, D. [1 ]
Li, K. [2 ]
机构
[1] Univ London Imperial Coll Sci Technol & Med, Dept Math, London SW7 2AZ, England
[2] Uppsala Univ, Dept Math, S-75106 Uppsala, Sweden
基金
英国工程与自然科学研究理事会;
关键词
Stochastic partial differential equation; Nonlinear filtering; Zakai equation; Particle filters; Gaussian mixtures; L-2-convergence; TARGET TRACKING;
D O I
10.1016/j.spa.2015.01.008
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Stochastic filtering is defined as the estimation of a partially observed dynamical system. Approximating the solution of the filtering problem with Gaussian mixtures has been a very popular method since the 1970s. Despite nearly fifty years of development, the existing work is based on the success of the numerical implementation and is not theoretically justified. This paper fills this gap and contains a rigorous analysis of a new Gaussian mixture approximation to the solution of the filtering problem. We deduce the L-2-convergence rate for the approximating system and show some numerical examples to test the new algorithm. (C) 2015 Elsevier B.V. All rights reserved.
引用
收藏
页码:2643 / 2673
页数:31
相关论文
共 36 条
[31]   RECURSIVE BAYESIAN ESTIMATION USING GAUSSIAN SUMS [J].
SORENSON, HW ;
ALSPACH, DL .
AUTOMATICA, 1971, 7 (04) :465-+
[32]  
Sun XT, 2004, P AMER CONTR CONF, P2098
[33]   GAUSSIAN SUM APPROACH TO PHASE AND FREQUENCY ESTIMATION [J].
TAM, PKS ;
MOORE, JB .
IEEE TRANSACTIONS ON COMMUNICATIONS, 1977, 25 (09) :935-942
[34]  
van der Merwe R, 2003, 2003 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOL VI, PROCEEDINGS, P701
[35]   Modeling and decoding motor cortical activity using a switching Kalman filter [J].
Wu, W ;
Black, MJ ;
Mumford, D ;
Gao, Y ;
Bienenstock, E ;
Donoghue, JP .
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 2004, 51 (06) :933-942
[36]   ON OPTIMAL FILTERING OF DIFFUSION PROCESSES [J].
ZAKAI, M .
ZEITSCHRIFT FUR WAHRSCHEINLICHKEITSTHEORIE UND VERWANDTE GEBIETE, 1969, 11 (03) :230-&