Urban intersections represent a complex environment for autonomous vehicles
with many sources of uncertainty. The vehicle must plan in a stochastic
environment with potentially rapid changes in driver behavior. Providing an
efficient strategy to navigate through urban intersections is a difficult task.
This paper frames the problem of navigating unsignalized intersections as a
partially observable Markov decision process (POMDP) and solves it using a
Monte Carlo sampling method. Empirical results in simulation show that the
resulting policy outperforms a threshold-based heuristic strategy on several
relevant metrics that measure both safety and efficiency.
%0 Generic
%1 bouton2017belief
%A Bouton, Maxime
%A Cosgun, Akansel
%A Kochenderfer, Mykel J.
%D 2017
%K learning reinforcement
%T Belief State Planning for Autonomously Navigating Urban Intersections
%U http://arxiv.org/abs/1704.04322
%X Urban intersections represent a complex environment for autonomous vehicles
with many sources of uncertainty. The vehicle must plan in a stochastic
environment with potentially rapid changes in driver behavior. Providing an
efficient strategy to navigate through urban intersections is a difficult task.
This paper frames the problem of navigating unsignalized intersections as a
partially observable Markov decision process (POMDP) and solves it using a
Monte Carlo sampling method. Empirical results in simulation show that the
resulting policy outperforms a threshold-based heuristic strategy on several
relevant metrics that measure both safety and efficiency.
@misc{bouton2017belief,
abstract = {Urban intersections represent a complex environment for autonomous vehicles
with many sources of uncertainty. The vehicle must plan in a stochastic
environment with potentially rapid changes in driver behavior. Providing an
efficient strategy to navigate through urban intersections is a difficult task.
This paper frames the problem of navigating unsignalized intersections as a
partially observable Markov decision process (POMDP) and solves it using a
Monte Carlo sampling method. Empirical results in simulation show that the
resulting policy outperforms a threshold-based heuristic strategy on several
relevant metrics that measure both safety and efficiency.},
added-at = {2017-04-18T07:52:25.000+0200},
author = {Bouton, Maxime and Cosgun, Akansel and Kochenderfer, Mykel J.},
biburl = {https://www.bibsonomy.org/bibtex/25a234ac7f26f38c570df1e8f29440b9a/nrlugg},
interhash = {ba30549fd2b5093e2e989a90327397a1},
intrahash = {5a234ac7f26f38c570df1e8f29440b9a},
keywords = {learning reinforcement},
note = {cite arxiv:1704.04322Comment: 6 pages, 6 figures, accepted to IV2017},
timestamp = {2017-04-18T07:52:25.000+0200},
title = {Belief State Planning for Autonomously Navigating Urban Intersections},
url = {http://arxiv.org/abs/1704.04322},
year = 2017
}