Congestion Characteristics And Real-time Queue Length Estimation On Urban Expressway

Traffic congestion on expressway recently has been a major transportation issue, particularly in large or medium-sized city areas in China. Expressway construction has been aimed at alleviating urban congestion; unfortunately the increasing demand leads to severe expressway congestion(especially during morning /afternoon peak hours), which seriously influences people’s living standard.Expressway bottlenecks are the major cause of congestion. Taking the recurrent congestion on urban expressway as the studied object, this dissertation in depth analyzed and studied the congestion-related characteristic and real time queue length(or queue tail location) estimation methods. This study is important for relieving recurrent congestions on expressway.Recurrent congestion-related characteristics including breakdown(i.e,queue onset time), capacity, queue discharge flow and queue density were investigated and analyzed in this research.(1)Theoretically, breakdown will occur when the traffic flow rate exceeds the specific capacity. However, it is unveiled that the breakdown flow rate varies with scenario and is random by observing the expresseay field data; thus the specific expressway capacity cannot continue to be defined with respect to breakdown conditions. This research proposed the breakdown identification method based on the sudden and sharp speed drop.(2)Also, in this research, it is well established on expressway that capacity drops happen during congestion and the two capacity estimation method based on least square method has been presented.(3)Queue discharge flow is another crucial factor for queue estimation and has been widely assumed as deterministic behavior in previous studies. To further improve the queue estimation performance on expressway, the queue discharge flow was treated as a time dependent variable in queue estimation model after carefully invesitigating the dynamic queue discharge flow using field data.(4)In addition, queue density, an indicator of traffic congestion, is closely related to queue length. This research proposed queue density estimation method considering capacity drop phenomenon from the linear congested portion of the density–flow relationship.This dissertation, based on the congestion-related characteristics, developped the time-space discrete macroscopic models respectively using the input-output theory and shockwave theory for real-time queue length(queue tail location) estimation on expressway.(1)Traditional input-output model, developed based on input-output theory, has been widely applied as a common technique for estimating queue length in the uninterrupted flow. Unfortunately, several researchers argued that the traditional input-output model underestimates queue length. The traditional input-output model was analyzed regarding with the reason of queue length underestimation and modified to develop a novel model for estimating real-time queue length on expressway.(2)As another common method for queue estimation, shockwave theory has been widely adopted for queue length estimation in an interrupted flow situation, rather than in an uninterrupted flow situation, such as on expressway. To bridge these gaps, this dissertation developed a discrete time-space model for real-time queue length estimation based on shockwave theory, i.e., shockwave-based model. The validation tasks of two proposed models were performed by VISSIM simulation platform; the validation results illustrated that both proposed queue estimation models are suitable for estimating expressway queue length.Over the past few decades, several active traffic control methods based on Intelligent Transportation Systems(ITS) have been developed and implemented to mitigate traffic congestion. Active Traffic Management(ATM) strategy is a Green and effective measure for queue congestion alleviation. Based on research achievements incongestion characteristics and queue length estimation models, this disseration discussed and advised how to develop ATM strategiesaimed at alleviating expressway congestion. Ramp metering(RM), route guidance(RG) and variable speed limits(VSL) are the most widely used ATM methods. Among them, RM is currentlytreated as the most widely practiced ATM method onabroad for relieving congestion on urban expressway or freeway. To counter the domestic expressway congestion, this dissertationsuggested that the queue length on mainline would be applied in developing thecoordinated RM stragetywhose framework wasalso given.RG by Variable Message Signs(VMS) is considered the most widely used control method for domestic expressway congestion. Taking the case of Erhuan Road in Xi’an, this research advanced a proposal for improving the traffic information exhibited on VMS by applying the proposed queue estimation model. Also, this research modified the existing Dyna TAM-VSL model aiming at relieving congestion, the effect of which on congestion was explored. Beyond that, this dissertation analyzed the incentives of a queue build-up in the VSL control situation. To effectively relieving expressway congestion, this research presented a variable speed limit control(VSL) strategy framework where real-time queue length would be taken in consideration.