A stage pressure-based adaptive traffic signal control using reinforcement learning

Fuyu Hu; Wei Huang

doi:10.1117/12.2627816

14 March 2022 A stage pressure-based adaptive traffic signal control using reinforcement learning

Fuyu Hu, Wei Huang

Proceedings Volume 12165, International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2021); 121651M (2022) https://doi.org/10.1117/12.2627816
Event: 2021 International Conference on Intelligent Traffic Systems and Smart City, 2021, Zhengzhou, China

Abstract

An adaptive traffic signal control model for isolated intersections is proposed based on reinforcement learning. The signal control problem is formulated as a Markov Decision Process because its probabilistic features match well with the random nature of traffic system. To improve the design of reward, which is usually an ad-hoc combination of several traffic measures, this paper develops a reinforcement learning algorithm that draws connections between the performance evaluation with the stage pressure-based P0 control policy in the reward design process. Numerical results show that the reinforcement learning control method has lower total delay and higher throughput compared with the fixed-time control. Moreover, the P0 policy performs better in improving throughput under heavy traffic conditions.

Citation Download Citation

Fuyu Hu and Wei Huang "A stage pressure-based adaptive traffic signal control using reinforcement learning", Proc. SPIE 12165, International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2021), 121651M (14 March 2022); https://doi.org/10.1117/12.2627816

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available