A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles
The advent of connected vehicle (CV) technology offers new possibilities for a revolution in future transportation systems. With the availability of real-time traffic data from CVs, it is possible to more effectively optimize traffic signals to reduce congestion, increase fuel efficiency, and enhanc...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/4/2750 |
| _version_ | 1797622520157831168 |
|---|---|
| author | Yang Shi Zhenbo Wang Tim J. LaClair Chieh (Ross) Wang Yunli Shao Jinghui Yuan |
| author_facet | Yang Shi Zhenbo Wang Tim J. LaClair Chieh (Ross) Wang Yunli Shao Jinghui Yuan |
| author_sort | Yang Shi |
| collection | DOAJ |
| description | The advent of connected vehicle (CV) technology offers new possibilities for a revolution in future transportation systems. With the availability of real-time traffic data from CVs, it is possible to more effectively optimize traffic signals to reduce congestion, increase fuel efficiency, and enhance road safety. The success of CV-based signal control depends on an accurate and computationally efficient model that accounts for the stochastic and nonlinear nature of the traffic flow. Without the necessity of prior knowledge of the traffic system’s model architecture, reinforcement learning (RL) is a promising tool to acquire the control policy through observing the transition of the traffic states. In this paper, we propose a novel data-driven traffic signal control method that leverages the latest in deep learning and reinforcement learning techniques. By incorporating a compressed representation of the traffic states, the proposed method overcomes the limitations of the existing methods in defining the action space to include more practical and flexible signal phases. The simulation results demonstrate the convergence and robust performance of the proposed method against several existing benchmark methods in terms of average vehicle speeds, queue length, wait time, and traffic density. |
| first_indexed | 2024-03-11T09:11:27Z |
| format | Article |
| id | doaj.art-e18cc53342f04942bfd52b94bb8d497a |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-11T09:11:27Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-e18cc53342f04942bfd52b94bb8d497a2023-11-16T19:00:12ZengMDPI AGApplied Sciences2076-34172023-02-01134275010.3390/app13042750A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected VehiclesYang Shi0Zhenbo Wang1Tim J. LaClair2Chieh (Ross) Wang3Yunli Shao4Jinghui Yuan5Department of Mechanical, Aerospace and Biomedical Engineering, University of Tennessee, Knoxville, TN 37996, USADepartment of Mechanical, Aerospace and Biomedical Engineering, University of Tennessee, Knoxville, TN 37996, USABuildings and Transportation Science Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USABuildings and Transportation Science Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USABuildings and Transportation Science Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USABuildings and Transportation Science Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USAThe advent of connected vehicle (CV) technology offers new possibilities for a revolution in future transportation systems. With the availability of real-time traffic data from CVs, it is possible to more effectively optimize traffic signals to reduce congestion, increase fuel efficiency, and enhance road safety. The success of CV-based signal control depends on an accurate and computationally efficient model that accounts for the stochastic and nonlinear nature of the traffic flow. Without the necessity of prior knowledge of the traffic system’s model architecture, reinforcement learning (RL) is a promising tool to acquire the control policy through observing the transition of the traffic states. In this paper, we propose a novel data-driven traffic signal control method that leverages the latest in deep learning and reinforcement learning techniques. By incorporating a compressed representation of the traffic states, the proposed method overcomes the limitations of the existing methods in defining the action space to include more practical and flexible signal phases. The simulation results demonstrate the convergence and robust performance of the proposed method against several existing benchmark methods in terms of average vehicle speeds, queue length, wait time, and traffic density.https://www.mdpi.com/2076-3417/13/4/2750traffic signal controldeep reinforcement learningautoencoder neural networkrepresentation learning |
| spellingShingle | Yang Shi Zhenbo Wang Tim J. LaClair Chieh (Ross) Wang Yunli Shao Jinghui Yuan A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles Applied Sciences traffic signal control deep reinforcement learning autoencoder neural network representation learning |
| title | A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles |
| title_full | A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles |
| title_fullStr | A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles |
| title_full_unstemmed | A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles |
| title_short | A Novel Deep Reinforcement Learning Approach to Traffic Signal Control with Connected Vehicles |
| title_sort | novel deep reinforcement learning approach to traffic signal control with connected vehicles |
| topic | traffic signal control deep reinforcement learning autoencoder neural network representation learning |
| url | https://www.mdpi.com/2076-3417/13/4/2750 |
| work_keys_str_mv | AT yangshi anoveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT zhenbowang anoveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT timjlaclair anoveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT chiehrosswang anoveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT yunlishao anoveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT jinghuiyuan anoveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT yangshi noveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT zhenbowang noveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT timjlaclair noveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT chiehrosswang noveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT yunlishao noveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles AT jinghuiyuan noveldeepreinforcementlearningapproachtotrafficsignalcontrolwithconnectedvehicles |