A novel particle weight optimization method based on multi-sensor observation fusion

Author

Chun-ling, Fu ; De-long, Gong ; Peiyan, Jia

Author_Institution

Basic Exp. Teaching Center, Henan Univ., Kaifeng, China

fYear

2012

fDate

25-27 July 2012

Firstpage

762

Lastpage

766

Abstract

Aiming at the effective realization of particle filter in multi-sensor observation system, a novel particle weight optimization method based on multi-sensor observation fusion is proposed in this paper. In the new algorithm, the observation likelihood function is firstly constructed on the basis of the concrete form of proposal distribution, and all observations at current sampling time are used to calculate particle weight, respectively. Next, on the basic assumption of sensors with identical accuracy, combined with average weighted strategy, the weighting fusion method is used to further optimize every particle weight in multi-sensor observation. Finally, the filter precision is improved by decreasing the variance of particle weights. The theoretical analysis and experimental results show the feasibility and efficiency of the proposed algorithm.

Keywords

optimisation; particle filtering (numerical methods); sensor fusion; signal sampling; statistical distributions; average weighted strategy; filter precision improvement; multisensor observation fusion; multisensor observation system; observation likelihood function; particle filter; particle weight optimization method; proposal distribution; sampling time; weighting fusion method; Filtering algorithms; Monte Carlo methods; Noise; Optimization methods; Particle filters; Sensors; Multi-sensor Observation; Nonlinear Estimation; Particle Filter; Weight Optimization;

fLanguage

English

Publisher

ieee

Conference_Titel

Control Conference (CCC), 2012 31st Chinese

Conference_Location

Hefei

ISSN

1934-1768

Print_ISBN

978-1-4673-2581-3

Type

conf

Filename

6390027