Traffic queuing on freeways brings travel delays, air pollution and waste of energy. Identifying traffic queues accurately and timely is a prerequisite for making strategies of alleviating traffic congestion. The occurrence of traffic queues on freeways is of randomness and uncertainty. This research proposed a methodology of real-time estimating traffic queues on freeways based on traffic data collected by environment sensors installed on instrumented vehicles. The proposed methodology is based on the mechanism of traffic shockwaves and the characteristics of instrumented vehicle data. The methodology consists of five steps: de-noising the instrumented vehicle data, identifying turning points, grouping turning points, estimating traffic shockwave speed and estimating queue length. One dataset from NGSIM is utilized to evaluate the performance of the proposed methodology. The results show that the methodology has a good performance in detecting the occurrence of traffic queues and estimating the length of queues. And the higher penetration rates of instrumented vehicle would bring more accurate estimation results. The proposed methodology has the advantages of real-time, high efficiency and accuracy in estimating queue on freeways.

CICTP