The queue length estimation for congested signalized intersections based on shockwave theory

Abstract

The queue length at signalized intersection is critical to either signal performance measure in terms of vehicle delay and stops or signal optimization. The traditional deterministic queuing theory based on the hypothesis that the flow of the road is steady during the entire time which disagrees with the actual situation. In this paper, instead of instead of using the deterministic queuing theory we solve the problem by modeling the queue dynamics in the signalized intersection with the Lighthill-Whitham-Richards(LWR) shockwave theory. We use the detector to identify traffic state changes by analysis the data in the immediate past cycle collected by the detector settled in the road. When the queue length is longer than the detector we can still distinguish the queue discharge flow from upstream arrival traffic. Therefore our approach can estimate queue length of the saturated Intersections. We used VISSIM to comparing our model with the SIGNAL94 model and evaluated by comparing the estimate maximum queue length with the queue length outputted by VISSIM. The results demonstrate that the proposed model can estimate long queues with satisfactory accuracy.