Traffic light dispersion control based on deep reinforcement learning
The current traffic light controls are ineffective and causes a handful of problems such as congestion and pollution. This study investigates the application of deep reinforcement learning on traffic control systems to minimize congestion at traffic intersection. The traffic data from Pulai Perdana,...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Penerbit UTM Press
2019
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/84965/1/Kamarulafizamismail2019_TrafficLightDispersionControl.pdf |
| _version_ | 1796863951396405248 |
|---|---|
| author | Bryan, Chua Ismail, Kamarulafizam Mohd. Zawawi, Fazila Mohd. Nor, Nur Safwati |
| author_facet | Bryan, Chua Ismail, Kamarulafizam Mohd. Zawawi, Fazila Mohd. Nor, Nur Safwati |
| author_sort | Bryan, Chua |
| collection | ePrints |
| description | The current traffic light controls are ineffective and causes a handful of problems such as congestion and pollution. This study investigates the application of deep reinforcement learning on traffic control systems to minimize congestion at traffic intersection. The traffic data from Pulai Perdana, Skudai, Johor Intersection was extracted, analysed and simulated based on the Poisson Distribution, using a simulator, Simulation of Urban Mobility (SUMO). In this research, we proposed a deep reinforcement learning model, which combines the capabilities of convolutional neural networks and reinforcement learning to control the traffic lights to increase the effectiveness of the traffic control system. The paper explains the method we used to quantify the traffic scenario into different matrices which fed to the model as states which reduces the load of computing as compared to images. After 2000 iterations of training, our deep reinforcement learning model was able to reduce the cumulative waiting time of all the vehicles at the Pulai Perdana intersection by 47.31% as compared to a fixed time algorithm and can perform even when the traffic is skewed in a different direction. When the traffic is scaled down to 50% and 20 %, the agent continues to improve the waiting time by 69.5% and 68.36 % respectively. It is proven in the experiment that a deep reinforcement learning model was able to reduce the cumulative waiting time at Pulai Perdana by 47.31%. |
| first_indexed | 2024-03-05T20:34:24Z |
| format | Article |
| id | utm.eprints-84965 |
| institution | Universiti Teknologi Malaysia - ePrints |
| language | English |
| last_indexed | 2024-03-05T20:34:24Z |
| publishDate | 2019 |
| publisher | Penerbit UTM Press |
| record_format | dspace |
| spelling | utm.eprints-849652020-02-29T13:59:57Z http://eprints.utm.my/84965/ Traffic light dispersion control based on deep reinforcement learning Bryan, Chua Ismail, Kamarulafizam Mohd. Zawawi, Fazila Mohd. Nor, Nur Safwati TJ Mechanical engineering and machinery The current traffic light controls are ineffective and causes a handful of problems such as congestion and pollution. This study investigates the application of deep reinforcement learning on traffic control systems to minimize congestion at traffic intersection. The traffic data from Pulai Perdana, Skudai, Johor Intersection was extracted, analysed and simulated based on the Poisson Distribution, using a simulator, Simulation of Urban Mobility (SUMO). In this research, we proposed a deep reinforcement learning model, which combines the capabilities of convolutional neural networks and reinforcement learning to control the traffic lights to increase the effectiveness of the traffic control system. The paper explains the method we used to quantify the traffic scenario into different matrices which fed to the model as states which reduces the load of computing as compared to images. After 2000 iterations of training, our deep reinforcement learning model was able to reduce the cumulative waiting time of all the vehicles at the Pulai Perdana intersection by 47.31% as compared to a fixed time algorithm and can perform even when the traffic is skewed in a different direction. When the traffic is scaled down to 50% and 20 %, the agent continues to improve the waiting time by 69.5% and 68.36 % respectively. It is proven in the experiment that a deep reinforcement learning model was able to reduce the cumulative waiting time at Pulai Perdana by 47.31%. Penerbit UTM Press 2019 Article PeerReviewed application/pdf en http://eprints.utm.my/84965/1/Kamarulafizamismail2019_TrafficLightDispersionControl.pdf Bryan, Chua and Ismail, Kamarulafizam and Mohd. Zawawi, Fazila and Mohd. Nor, Nur Safwati (2019) Traffic light dispersion control based on deep reinforcement learning. Journal of Transport System Engineering, 6 (1). pp. 45-53. ISSN 2289–9790 https://jtse.utm.my/index.php/jtse/article/download/97/100 |
| spellingShingle | TJ Mechanical engineering and machinery Bryan, Chua Ismail, Kamarulafizam Mohd. Zawawi, Fazila Mohd. Nor, Nur Safwati Traffic light dispersion control based on deep reinforcement learning |
| title | Traffic light dispersion control based on deep reinforcement learning |
| title_full | Traffic light dispersion control based on deep reinforcement learning |
| title_fullStr | Traffic light dispersion control based on deep reinforcement learning |
| title_full_unstemmed | Traffic light dispersion control based on deep reinforcement learning |
| title_short | Traffic light dispersion control based on deep reinforcement learning |
| title_sort | traffic light dispersion control based on deep reinforcement learning |
| topic | TJ Mechanical engineering and machinery |
| url | http://eprints.utm.my/84965/1/Kamarulafizamismail2019_TrafficLightDispersionControl.pdf |
| work_keys_str_mv | AT bryanchua trafficlightdispersioncontrolbasedondeepreinforcementlearning AT ismailkamarulafizam trafficlightdispersioncontrolbasedondeepreinforcementlearning AT mohdzawawifazila trafficlightdispersioncontrolbasedondeepreinforcementlearning AT mohdnornursafwati trafficlightdispersioncontrolbasedondeepreinforcementlearning |