Analytically Guided Reinforcement Learning for Green It and Fluent Traffic
This study investigates various methods for autonomous traffic signal control. We look into different types of control methods, including fixed time, adaptive, analytic, and reinforcement learning approaches. Machine learning approaches are compared with the “analytic” approach...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2022-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9875292/ |
_version_ | 1797995414779068416 |
---|---|
author | Marcin Korecki Dirk Helbing |
author_facet | Marcin Korecki Dirk Helbing |
author_sort | Marcin Korecki |
collection | DOAJ |
description | This study investigates various methods for autonomous traffic signal control. We look into different types of control methods, including fixed time, adaptive, analytic, and reinforcement learning approaches. Machine learning approaches are compared with the “analytic” approach, which is used as “gold standard” for performance assessment. We find that conventional machine learning approaches are better than the analytic approach, but require a lot more computer power. We, therefore, introduce a novel hybrid method called “analytically guided reinforcement learning” or shorter “<inline-formula> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula>-RL”. This approach is implemented in our “GuidedLight agent” and tends to outperform both, classical machine learning and the analytic approach, while largely improving convergence. This method is therefore suited as a “green IT” solution that improves environmental impact in a two-fold way: by reducing (i) traffic congestion and (ii) the processing power needed for the learning and operation of the traffic light control algorithm. |
first_indexed | 2024-04-11T10:01:14Z |
format | Article |
id | doaj.art-f49df5bcfdcd4104b28075f5656102f4 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-04-11T10:01:14Z |
publishDate | 2022-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-f49df5bcfdcd4104b28075f5656102f42022-12-22T04:30:26ZengIEEEIEEE Access2169-35362022-01-0110963489635810.1109/ACCESS.2022.32040579875292Analytically Guided Reinforcement Learning for Green It and Fluent TrafficMarcin Korecki0Dirk Helbing1ETH Zürich, Computational Social Science, ETH Zürich, Zürich, SwitzerlandETH Zürich, Computational Social Science, ETH Zürich, Zürich, SwitzerlandThis study investigates various methods for autonomous traffic signal control. We look into different types of control methods, including fixed time, adaptive, analytic, and reinforcement learning approaches. Machine learning approaches are compared with the “analytic” approach, which is used as “gold standard” for performance assessment. We find that conventional machine learning approaches are better than the analytic approach, but require a lot more computer power. We, therefore, introduce a novel hybrid method called “analytically guided reinforcement learning” or shorter “<inline-formula> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula>-RL”. This approach is implemented in our “GuidedLight agent” and tends to outperform both, classical machine learning and the analytic approach, while largely improving convergence. This method is therefore suited as a “green IT” solution that improves environmental impact in a two-fold way: by reducing (i) traffic congestion and (ii) the processing power needed for the learning and operation of the traffic light control algorithm.https://ieeexplore.ieee.org/document/9875292/Green AIcomplex systemsreinforcement learningsmart citiessustainabilitytraffic light control |
spellingShingle | Marcin Korecki Dirk Helbing Analytically Guided Reinforcement Learning for Green It and Fluent Traffic IEEE Access Green AI complex systems reinforcement learning smart cities sustainability traffic light control |
title | Analytically Guided Reinforcement Learning for Green It and Fluent Traffic |
title_full | Analytically Guided Reinforcement Learning for Green It and Fluent Traffic |
title_fullStr | Analytically Guided Reinforcement Learning for Green It and Fluent Traffic |
title_full_unstemmed | Analytically Guided Reinforcement Learning for Green It and Fluent Traffic |
title_short | Analytically Guided Reinforcement Learning for Green It and Fluent Traffic |
title_sort | analytically guided reinforcement learning for green it and fluent traffic |
topic | Green AI complex systems reinforcement learning smart cities sustainability traffic light control |
url | https://ieeexplore.ieee.org/document/9875292/ |
work_keys_str_mv | AT marcinkorecki analyticallyguidedreinforcementlearningforgreenitandfluenttraffic AT dirkhelbing analyticallyguidedreinforcementlearningforgreenitandfluenttraffic |