The Unscented Particle Filter

DMCA

The Unscented Particle Filter (2000)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Rudolph van der Merwe , Arnaud Doucet , Nando de Freitas , Eric Wan

Citations:

211 - 8 self

BibTeX

@MISC{Merwe00theunscented,
    author = {Rudolph van der Merwe and Arnaud Doucet and Nando de Freitas and Eric Wan},
    title = {The Unscented Particle Filter},
    year = {2000}
}

OpenURL

Abstract

In this paper, we propose a new particle filter based on sequential importance sampling. The algorithm uses a bank of unscented filters to obtain the importance proposal distribution. This proposal has two very "nice" properties. Firstly, it makes efficient use of the latest available information and, secondly, it can have heavy tails. As a result, we find that the algorithm outperforms standard particle filtering and other nonlinear filtering methods very substantially. This experimental finding is in agreement with the theoretical convergence proof for the algorithm. The algorithm also includes resampling and (possibly) Markov chain Monte Carlo (MCMC) steps.

Keyphrases

unscented particle filter theoretical convergence proof nice property unscented filter new particle filter available information efficient use sequential importance sampling heavy tail standard particle filtering experimental finding importance proposal distribution markov chain monte carlo

DMCA