DocumentCode :
477964
Title :
A New Particle Filter and Its Application in Mobile Robot Localization
Author :
Xia, Yi-Min ; Yang, Yi-Min
Author_Institution :
Acad. of Autom., Guangdong Univ. of Technol., Guangzhou
Volume :
4
fYear :
2008
fDate :
18-20 Oct. 2008
Firstpage :
522
Lastpage :
525
Abstract :
Particle filter (PF) is widely used in mobile robot localization, since it is suitable for nonlinear non-Gaussian system. In order to get rid of the bug that the performance of traditional PF is seriously dependent on the selection of proposal distribution, we put forward a unscented particle filter (UPF) algorithm by importing the unscented Kalman filter (UKF) to generate the proposal distribution, which means using a series of confirmed samples to approximate the posterior probability density function of the state. Thus the generated proposal distribution will approximate the real posterior probability density function much better, and the quality of the traditional PF will get improved. The simulation result shows that the performance of improved algorithm is better than the traditional particle filter although the run time is longer.
Keywords :
Kalman filters; approximation theory; mobile robots; particle filtering (numerical methods); probability; mobile robot localization; nonlinear nonGaussian system; posterior probability density function approximation; unscented Kalman filter; unscented particle filter algorithm; Distribution functions; Fuzzy systems; Gaussian noise; Mobile robots; Monte Carlo methods; Particle filters; Probability density function; Proposals; Robotics and automation; Sampling methods; Particle Filter; Unscented Kalman Filter; location; mobile robot;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Fuzzy Systems and Knowledge Discovery, 2008. FSKD '08. Fifth International Conference on
Conference_Location :
Jinan Shandong
Print_ISBN :
978-0-7695-3305-6
Type :
conf
DOI :
10.1109/FSKD.2008.133
Filename :
4666440
Link To Document :
بازگشت