Reinforcement Learning Ramp Metering without Complete Information

This paper develops a model of reinforcement learning ramp metering (RLRM) without complete information, which is applied to alleviate traffic congestions on ramps. RLRM consists of prediction tools depending on traffic flow simulation and optimal choice model based on reinforcement learning theorie...

Full description

Saved in:
Bibliographic Details
Main Authors: Xing-Ju Wang, Xiao-Ming Xi, Gui-Feng Gao
Format: Article
Language:English
Published: Wiley 2012-01-01
Series:Journal of Control Science and Engineering
Online Access:http://dx.doi.org/10.1155/2012/208456
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849472921881280512
author Xing-Ju Wang
Xiao-Ming Xi
Gui-Feng Gao
author_facet Xing-Ju Wang
Xiao-Ming Xi
Gui-Feng Gao
author_sort Xing-Ju Wang
collection DOAJ
description This paper develops a model of reinforcement learning ramp metering (RLRM) without complete information, which is applied to alleviate traffic congestions on ramps. RLRM consists of prediction tools depending on traffic flow simulation and optimal choice model based on reinforcement learning theories. Moreover, it is also a dynamic process with abilities of automaticity, memory and performance feedback. Numerical cases are given in this study to demonstrate RLRM such as calculating outflow rate, density, average speed, and travel time compared to no control and fixed-time control. Results indicate that the greater is the inflow, the more is the effect. In addition, the stability of RLRM is better than fixed-time control.
format Article
id doaj-art-db226ffe45944a6488a67cd4d85990c2
institution Kabale University
issn 1687-5249
1687-5257
language English
publishDate 2012-01-01
publisher Wiley
record_format Article
series Journal of Control Science and Engineering
spelling doaj-art-db226ffe45944a6488a67cd4d85990c22025-08-20T03:24:21ZengWileyJournal of Control Science and Engineering1687-52491687-52572012-01-01201210.1155/2012/208456208456Reinforcement Learning Ramp Metering without Complete InformationXing-Ju Wang0Xiao-Ming Xi1Gui-Feng Gao2School of Traffic and Transportation, Shijiazhuang Tiedao University, Shijiazhuang, Hebei 050043, ChinaSchool of Traffic and Transportation, Shijiazhuang Tiedao University, Shijiazhuang, Hebei 050043, ChinaSchool of Traffic and Transportation, Shijiazhuang Tiedao University, Shijiazhuang, Hebei 050043, ChinaThis paper develops a model of reinforcement learning ramp metering (RLRM) without complete information, which is applied to alleviate traffic congestions on ramps. RLRM consists of prediction tools depending on traffic flow simulation and optimal choice model based on reinforcement learning theories. Moreover, it is also a dynamic process with abilities of automaticity, memory and performance feedback. Numerical cases are given in this study to demonstrate RLRM such as calculating outflow rate, density, average speed, and travel time compared to no control and fixed-time control. Results indicate that the greater is the inflow, the more is the effect. In addition, the stability of RLRM is better than fixed-time control.http://dx.doi.org/10.1155/2012/208456
spellingShingle Xing-Ju Wang
Xiao-Ming Xi
Gui-Feng Gao
Reinforcement Learning Ramp Metering without Complete Information
Journal of Control Science and Engineering
title Reinforcement Learning Ramp Metering without Complete Information
title_full Reinforcement Learning Ramp Metering without Complete Information
title_fullStr Reinforcement Learning Ramp Metering without Complete Information
title_full_unstemmed Reinforcement Learning Ramp Metering without Complete Information
title_short Reinforcement Learning Ramp Metering without Complete Information
title_sort reinforcement learning ramp metering without complete information
url http://dx.doi.org/10.1155/2012/208456
work_keys_str_mv AT xingjuwang reinforcementlearningrampmeteringwithoutcompleteinformation
AT xiaomingxi reinforcementlearningrampmeteringwithoutcompleteinformation
AT guifenggao reinforcementlearningrampmeteringwithoutcompleteinformation