A Fast Method to Prevent Traffic Blockage by Signal Control Based on Reinforcement Learning

DOI

10.2991/ceie-16.2017.36How to use a DOI?

Keywords

Q-Learning Algorithm; Traffic Modeling; Wavelet Neural Network; Traffic Signal Control

Abstract

In this paper, we present an efficient and fast way to prevent traffic blockage by controlling traffic signal. A new model is adopted to out program which is fused and developed by probabilistic model and cellular automatic model (CA model). Based on this model, we used wavelet neural network (WNN) for predicting traffic flow and use this data to improve the green or red light time sitting. Q-learning algorithm, as one of the reinforcement learning methods, also is applied to project the phase of traffic light from different intersections. Considering the requirement of model response rate, we also develop this algorithm for a fast respond to the complex and fickle traffic condition. Finally, a simulation study is carried out to evaluate out method and the result shows that this method can avoid traffic blockage efficiently.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Mengjia Shen
PY  - 2016/10
DA  - 2016/10
TI  - A Fast Method to Prevent Traffic Blockage by Signal Control Based on Reinforcement Learning
BT  - Proceedings of the International Conference on Communication and Electronic Information Engineering (CEIE 2016)
PB  - Atlantis Press
SP  - 284
EP  - 291
SN  - 2352-5401
UR  - https://doi.org/10.2991/ceie-16.2017.36
DO  - 10.2991/ceie-16.2017.36
ID  - Shen2016/10
ER  -