Fig 1.
A road network composing of roads and intersections.
Table 1.
Attributes of the road entities.
Table 2.
Attributes of the intersection entities.
Fig 2.
The “lane states” of several vehicles driving on two linked roads.
Note that the right-most vehicle is occupying lanes with indices 1 and 2 from its point of view, but the latter is in fact lane 1 on the linked road.
Table 3.
Attributes of the vehicle entities related to spatial location.
Table 4.
Attributes of the vehicle entities related to path initialization.
Table 5.
Attributes related to speed computation.
Table 6.
Attributes related to lane change.
Table 7.
Selection of candidate lane parameters.
Table 8.
Driver parameter values for the circuit.
Fig 3.
Average speed value according to the number of cars.
Table 9.
Computation time according to the number of vehicles.
Fig 4.
Tay Son street site (based on OpenStreetMap data): The green circle represents the input point (WGS84 coordinates: (105.8224 21.005)), and the red circle the output (WGS84 coordinates: (105.8236 21.0076)).
Fig 5.
Chua Boc street site (based on OpenStreetMap data): The green circle represents the input point (WGS84 coordinates: (105.8252 21.0089)), and the red circle the output point (WGS84 coordinates: (105.831 21.0062)).
Table 10.
Characteristic of the 2 sites.
Fig 6.
Count of vehicles at the input and output points for the Tay Son site (time series were smoothed out using moving average with a window size of 10s).
Fig 7.
Count of vehicles at the input and output points for the Chua Boc site (time series were smoothed out using moving average with a window size of 10s).
Table 11.
Driver parameter values for Hanoi.
Fig 8.
Simulated results for the Tay Son site.
Motorcycle and car counts for the 100 simulations with [44] model (time series were smoothed out using moving average with a window size of 10s). In red, the mean values.
Fig 9.
Simulated results for the Tay Son site.
Motorcycle and car counts for the 100 simulations with our model (time series were smoothed out using moving average with a window size of 10s). In red, the mean values.
Fig 10.
Observed and simulated results (with [44] and our models) for the Tay Son site.
Motorcycle and car counts—mean of the simulation (time series were smoothed out using moving average with a window size of 10s.
Table 12.
Mean metrics computed for the Tay Son site.
In parenthesis, the standard deviation.
Fig 11.
Simulated results for the Chua Boc site.
Motorcycle and car counts for the 100 simulations with [44] model (time series were smoothed out using moving average with a window size of 10s). In red, the mean values.
Fig 12.
Simulated results for the Chua Boc site.
Motorcycle and cars counts for the 100 simulations with our model (time series were smoothed out using moving average with a window size of 10s). In red, the mean values.
Fig 13.
Observed and simulated results (with [44] and our models) for the Chua Boc site.
Motorcycle and car counts—mean of the simulation (time series were smoothed out using moving average with a window size of 10s.
Table 13.
Mean metrics computed for the Chua Boc site.
In parenthesis, the standard deviation.