Topography-Aware Deep Reinforcement Learning with Contextual Reward Engineering for Sustainable and Efficient Urban Traffic Control
Future Transportation, Vol. 6, No. 2, Article 82
Editorial summary
Traffic-signal control is trained around road slope and vehicle mix. A simulation calibrated with traffic video compares emissions, journey time and queues on an uphill approach. The specialised controller performs better than pressure-based and standard deep Q-learning controls in this setting.
Bibliographic reference
Use the publisher’s record for the citation format required by your journal or organisation.
Oleksander Ryzhanskyi, Oleksander Barmak, Eduard Manziuk, Pavlo Radiuk, Iurii Krak. "Topography-Aware Deep Reinforcement Learning with Contextual Reward Engineering for Sustainable and Efficient Urban Traffic Control". Future Transportation, Vol. 6, No. 2, Article 82, 2026. https://doi.org/10.3390/futuretransp6020082
Test a method against your business task.
Describe the decision your system needs to support and the data you have. We assess whether a method fits, define a prototype and agree how to measure its quality, response time and operating cost.