S. Thrun, D. Fox, W. Burgard and F. Dellaert
Robust Monte Carlo localization
for mobile robots.
Artificial Intelligence
Abstract
Mobile robot localization is the problem of determining a robot's pose from
sensor data. This article presents a family of probabilisticlocalizationalgorithms
known as Monte Carlo Localization (MCL). MCL algorithms representa robot's
belief by a set of weightedhypotheses (samples), which approximatethe posterior
under a common Bayesian formulation of the localization problem.Building
onthe basic MCL algorithm, this article develops a more robust algorithmcalled
Mixture-MCL, which integrates two complimentary waysof generatingsamples
in the estimation. To apply this algorithm to mobile robots equippedwith
range finders, a kernel density tree islearned that permits fast sampling.
Systematic empirical results illustrate the robustness and computationalefficiency
of the approach.
Download
Full paper
[.pdf](1612 kb)
Bibtex
@Article{Thrun01Robust,
author = {Thrun, S. and Fox, D. and Burgard, W. and Dellaert.
F.},
title = {Robust Monte Carlo Localization for Mobile Robots},
journal = {Artificial Intelligence},
year = 2001,
volume = {128},
number = {1-2},
}