Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data
Reinforcement learning (RL)-based traffic signal control has been proven to have great potential in alleviating traffic congestion. The state definition, which is a key element in RL-based traffic signal control, plays a vital role. However, the data used for state definition in the literature are e...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-07-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/21/8/744 |
| _version_ | 1797999424035618816 |
|---|---|
| author | Song Wang Xu Xie Kedi Huang Junjie Zeng Zimin Cai |
| author_facet | Song Wang Xu Xie Kedi Huang Junjie Zeng Zimin Cai |
| author_sort | Song Wang |
| collection | DOAJ |
| description | Reinforcement learning (RL)-based traffic signal control has been proven to have great potential in alleviating traffic congestion. The state definition, which is a key element in RL-based traffic signal control, plays a vital role. However, the data used for state definition in the literature are either coarse or difficult to measure directly using the prevailing detection systems for signal control. This paper proposes a deep reinforcement learning-based traffic signal control method which uses high-resolution event-based data, aiming to achieve cost-effective and efficient adaptive traffic signal control. High-resolution event-based data, which records the time when each vehicle-detector actuation/de-actuation event occurs, is informative and can be collected directly from vehicle-actuated detectors (e.g., inductive loops) with current technologies. Given the event-based data, deep learning techniques are employed to automatically extract useful features for traffic signal control. The proposed method is benchmarked with two commonly used traffic signal control strategies, i.e., the fixed-time control strategy and the actuated control strategy, and experimental results reveal that the proposed method significantly outperforms the commonly used control strategies. |
| first_indexed | 2024-04-11T11:04:19Z |
| format | Article |
| id | doaj.art-a784eaa2df174b2e84f7b1517ba77b25 |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-04-11T11:04:19Z |
| publishDate | 2019-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-a784eaa2df174b2e84f7b1517ba77b252022-12-22T04:28:25ZengMDPI AGEntropy1099-43002019-07-0121874410.3390/e21080744e21080744Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based DataSong Wang0Xu Xie1Kedi Huang2Junjie Zeng3Zimin Cai4College of Systems Engineering, National University of Defense Technology, Changsha 410073, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha 410073, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha 410073, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha 410073, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha 410073, ChinaReinforcement learning (RL)-based traffic signal control has been proven to have great potential in alleviating traffic congestion. The state definition, which is a key element in RL-based traffic signal control, plays a vital role. However, the data used for state definition in the literature are either coarse or difficult to measure directly using the prevailing detection systems for signal control. This paper proposes a deep reinforcement learning-based traffic signal control method which uses high-resolution event-based data, aiming to achieve cost-effective and efficient adaptive traffic signal control. High-resolution event-based data, which records the time when each vehicle-detector actuation/de-actuation event occurs, is informative and can be collected directly from vehicle-actuated detectors (e.g., inductive loops) with current technologies. Given the event-based data, deep learning techniques are employed to automatically extract useful features for traffic signal control. The proposed method is benchmarked with two commonly used traffic signal control strategies, i.e., the fixed-time control strategy and the actuated control strategy, and experimental results reveal that the proposed method significantly outperforms the commonly used control strategies.https://www.mdpi.com/1099-4300/21/8/744traffic signal controldeep reinforcement learninghigh-resolution dataevent-based datadouble dueling deep Q network |
| spellingShingle | Song Wang Xu Xie Kedi Huang Junjie Zeng Zimin Cai Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data Entropy traffic signal control deep reinforcement learning high-resolution data event-based data double dueling deep Q network |
| title | Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data |
| title_full | Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data |
| title_fullStr | Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data |
| title_full_unstemmed | Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data |
| title_short | Deep Reinforcement Learning-Based Traffic Signal Control Using High-Resolution Event-Based Data |
| title_sort | deep reinforcement learning based traffic signal control using high resolution event based data |
| topic | traffic signal control deep reinforcement learning high-resolution data event-based data double dueling deep Q network |
| url | https://www.mdpi.com/1099-4300/21/8/744 |
| work_keys_str_mv | AT songwang deepreinforcementlearningbasedtrafficsignalcontrolusinghighresolutioneventbaseddata AT xuxie deepreinforcementlearningbasedtrafficsignalcontrolusinghighresolutioneventbaseddata AT kedihuang deepreinforcementlearningbasedtrafficsignalcontrolusinghighresolutioneventbaseddata AT junjiezeng deepreinforcementlearningbasedtrafficsignalcontrolusinghighresolutioneventbaseddata AT zimincai deepreinforcementlearningbasedtrafficsignalcontrolusinghighresolutioneventbaseddata |