LN Alegre, T Ziemke… - IEEE Transactions on …, 2021 - ieeexplore.ieee.org
… realistic scenario with multiple signalized intersections, this study considers a scenario of … with different traffic saturation conditions, we run the Cottbus scenario with different capacity …
… traffic signal control for different trafficscenarios. The proposed multi-agent reinforcement learning (… delay and speed in comparison to other traffic control system like hierarchical multi-…
J Ault, G Sharon - Thirty-fifth Conference on Neural Information …, 2021 - openreview.net
… We propose a toolkit for developing and comparing reinforcementlearning (RL)based traffic … control problems that are based on realistic trafficscenarios. Importantly, the toolkit allows a …
E Walraven, MTJ Spaan, B Bakker - Engineering Applications of Artificial …, 2016 - Elsevier
… We generate 20 policies for each trafficscenario and we run 5000 learning episodes. For each policy, the number of vehicle hours is computed and the results are shown in Fig. 8. …
X Liang, X Du, G Wang, Z Han - IEEE Transactions on …, 2019 - ieeexplore.ieee.org
… In this paper, we study how to decide the traffic signal … learning model to control the traffic light cycle. In the model, we quantify the complex trafficscenario as states by collecting traffic …
X Gao, X Li, Q Liu, Z Li, F Yang, T Luan - Sensors, 2022 - mdpi.com
… , which may reduce total traffic efficiency and the occurrence of traffic accidents. Graph … trafficscenarios. Therefore, some researchers focused on the graph reinforcementlearning (GRL) …
… To promote similar advances in traffic control via RL, we propose four … trafficscenarios, illustrating distinct reinforcement learning problems with applications to mixed-autonomy traffic. …
… various safe reinforcementlearning and multi-agent reinforcementlearning algorithms in … In five typical trafficscenarios such as roundabout and intersection, we study the problem of …
… traffic participants in dense merging scenarios. We show that deep reinforcementlearning policies can capture interaction patterns when trained in a variety of different scenarios, even …