Fig 1.
Fundamental diagram of traffic flow.
We show the relation between flow and density values obtained from computational simulations. The values correspond to the average flow obtained from 1000 simulations. At lower densities, we have the freeflow phase. The higher the density on the congested phase (negative slope), the lower the flow. The maximum flow corresponds to the capacity of the lane.
Fig 2.
Six-lane road with adaptive reversible lanes.
(A) 3-3 lane configuration. (B) 4-2 lane configuration. (C) 5-1 lane configuration. (D) 2-4 lane configuration. (E) 1-5 lane configuration.
Fig 3.
Heatmap of the traffic flow for all the lane configurations.
We present the traffic flow in terms of vehicles’ density and the percentage of vehicles to the east. Higher values of traffic flow correspond to the bright yellow color. We observe how the maximum traffic flow peak changes accordingly with the lane configuration. (A)5-1 lane configuration. (B)4-2 lane configuration, (C)3-3 lane configuration. (D)2-4 lane configuration. (E)1-5 lane configuration.
Fig 4.
Mean flow by lane configuration.
Mean flow as a function of the percentage of vehicles to the east direction for each lane configuration.
Fig 5.
Distribution of asymmetry values of traffic volumes for a loop detector located in Dundalk.
(A)Distribution of average asymmetry during 2013. (B)Distribution of average asymmetry in July 2013. (C)Distribution of average asymmetry between 15 and 19 of July 2013. (D)Distribution of asymmetry on July 17, 2013.
Fig 6.
Boxplot of the traffic asymmetry for a loop detector located in Dundalk in 2013.
We observe a high variance in the traffic volume during morning rush hours.
Fig 7.
Mean flow for the three scenarios.
In this figure, we show the mean flow for each scenario for the four locations. The adaptive lane scenario outperforms the reversible scenario. (A)Raheen 2012. (B)Rossbrien 2011. (C)Dunkettle 2007. (D)Rossbrien 2012.
Fig 8.
Configuration change ratio and traffic flow gain for the top 100 sensors.
We show the flow gain for the top 100 sensors, which have a higher configuration change ratio. We can see a relation between higher values of flow gain and configuration change ratio.