Accurate Traffic Flow Estimation in Urban Roads with Considering the Traffic Signals

摘要

The traffic condition can be improved with the real-time traffic information that is obtained from vehicle detectors (VDs) or probe vehicles (PVs). Using PVs has a lower cost and a broader coverage, but cannot measure the traffic flow like using VDs. Most studies on PVs used Fundamental Diagram (FD) models to investigate the speed-density-flow relationship. However, they didn't notice that the driving speed is varied with the traffic signal. Accordingly, we propose an approach, Flow Estimation with Traffic Signal (FETS), to estimate the traffic flow in urban roads by considering the traffic signal. The speed is calculated at green light and the density is acquired by the queue length at red light. The experiment results show that the mean relative error of FETS is 44.4% while the best one of the FD models is 117.3%, representing that FETS has better accuracy than FD models in urban roads.