Trafc Simulation Model SetupTo evaluate the impact of the CS approach on travel time estimation, a simulation model is built for a fve-mile twolane freeway segment using SUMO (Krajzewicz et al. 2012). The layout of the freeway segment is shown in Fig. 5.The fve-mile freeway segment consists of ten small segments, s1 to s10, of length 0.5 miles each. One loop detector and one RSU are located at every 0.5 mile. The total simulation time is 3600 s, and the trafc demand is set at 2400 vehicles. The traditional, non-connected vehicles (blue) and CVs (green) are generated based on a predefned CV penetration rate. The simulation period is split into 12 time intervals of 300 s, t1, …, t12. The frst three time intervals are considered as the warm-up period, and thus excluded from the analysis. In time intervals t7, t8 and t9, we close the inner lanes of segments s5 and s6 to create congestion conditions. Additional assumptions are as follows:
Trafc Simulation Model Setup<br>To evaluate the impact of the CS approach on travel time <br>estimation, a simulation model is built for a fve-mile twolane freeway segment using SUMO (Krajzewicz et al. 2012). <br>The layout of the freeway segment is shown in Fig. 5.<br>The fve-mile freeway segment consists of ten small segments, s1 to s10, of length 0.5 miles each. One loop detector <br>and one RSU are located at every 0.5 mile. The total simulation time is 3600 s, and the trafc demand is set at 2400 <br>vehicles. The traditional, non-connected vehicles (blue) <br>and CVs (green) are generated based on a predefned CV <br>penetration rate. The simulation period is split into 12 time <br>intervals of 300 s, t1, …, t12. The frst three time intervals are <br>considered as the warm-up period, and thus excluded from <br>the analysis. In time intervals t7, t8 and t9, we close the inner <br>lanes of segments s5 and s6 to create congestion conditions. <br>Additional assumptions are as follows:
正在翻译中..